Paris partner Pierre-Emmanuel Fender is the author of “Le rôle du fiduciaire dans une opération de fiducie,” [PDF] published in the n°154 issue of the French publication Revue LAMY Droit des Affaire in December 2019.

Brussels partner Peter Alexiadis and of counsel Alejandro Guerrero are the authors of “Sustainability Priorities and Competition Law Policies – A Meeting of Minds,” [PDF] published in Our World magazine in February 2020.

The EU has adopted a package of measures which will implement some important changes to the way in which alternative investment fund managers (“AIFMs”) market their funds cross-border in the EEA. One of the principal aims of the changes is to remove barriers to cross-border marketing of alternative investment funds (“AIFs”) which have arisen in large part due to differing implementation of the Alternative Investment Fund Managers Directive (“AIFMD”) across member states of the EEA since 2014.

The new definition of and conditions for “pre-marketing” will establish a defined phase of “pre-marketing” as distinct from “marketing” under the AIFMD. Importantly, once an EEA AIFM has engaged in “pre-marketing” activity in a jurisdiction any investment from an investor in that jurisdiction within 18-months of such pre-marketing will be deemed to be the result of marketing activity. Consequently, the ability to rely upon reverse solicitation will be curtailed within the EEA. The impact for non-EEA AIFMs is not yet clear since the CBD/R (as defined below) address only EEA AIFMs. However, it is highly unlikely that jurisdictions in the EEA will want to retain different pre-marketing regimes for EEA versus non-EEA AIFMs. This is something that non-EEA AIFMs looking to raise capital within the EEA in 2021 and beyond should keep under review.

The package of reforms is comprised of Directive (EU) 2019/1160 regarding the cross-border distribution of collective investment undertakings (“CBD”) and Regulation (EU) 2019/1156 on facilitating cross-border distribution of collective investment undertakings (“CBR”) (together “CBD/R”) and will come into force within the EEA on 2 August 2021.

The CBD/R apply to EEA AIFMs. However, it should be noted that Member States do not have a completely free hand in relation to their national private placement regimes (“NPPR”). It is highly likely that at least some of the changes will also impact how non-EEA AIFMs are able to market their funds into the EEA under the applicable NPPR. The full impact for non-EEA AIFMs will only be seen over time. However, non-EEA AIFMs should have the CBD/R on their radar and monitor changes to how they market their funds into the EEA as the implementation date approaches.

It should be noted that the CBD/R applies to the distribution of funds under both the AIFMD and the UCITS Directive. However, this briefing note focusses exclusively on the changes relevant to AIFMs.

Principal changes:

Harmonised definition and conditions of “pre-marketing”
Notification requirement to the national competent authority (“NCA”) of the AIFM when “pre-marketing” has commenced
Any subscription within 18-months of “pre-marketing” will be deemed to be the result of marketing
Any third party marketing an AIF on behalf of an AIFM must be authorised as a MiFID investment firm, credit institution, AIFM or UCITS management company
Restrictions on marketing of a successor fund following de-notification of marketing

Harmonised “pre-marketing” regime for EEA AIFMs

Since 2014, due to differences in the implementation of the AIFMD as regards when “marketing” commences for the purposes of the AIFMD, frustrations have been voiced regarding the inconsistent ability to pre-market across Europe in order to test the investor appetite for an investment strategy in the EEA. This arises because in some jurisdictions “marketing” is considered to occur at an early stage in the investment process, whereas in other jurisdictions (e.g. Luxembourg (and the UK)) “marketing” is considered to occur at a much later stage when the offer of interests in an AIF is capable of being accepted by investors on the basis of final form subscription documents.

The CBD/R introduces an EEA-wide definition of “pre-marketing”, i.e. those promotional activities which can take place before the marketing passport (which only applies once “marketing” has commenced) is available. The definition of “pre-marketing” is broad:

“provision of information or communication, direct or indirect, on investment strategies or investment ideas by an EU AIFM or on its behalf, to potential professional investors domiciled or with a registered office in the Union in order to test their interest in an AIF or a compartment which is not yet established, or which is established, but not yet notified for marketing in accordance with Article 31 or 32, in that Member State where the potential investors are domiciled or have their registered office, and which in each case does not amount to an offer or placement to the potential investor to invest in the units or shares of that AIF or compartment”.

Key features of “pre-marketing”:

Provision of information or communication on investment strategies or ideas (whether direct or indirect)
By an EEA AIFM or on its behalf (e.g. placement agent or non-EEA sponsor)
To potential professional investors in the EEA
In order to test their interest:
In an AIF/compartment which is not yet established; or
In an AIF/compartment which is established, but not yet notified for marketing under AIFMD in that Member State
In each case does not amount to an offer or placement to the potential investor to invest in the units or shares of that AIF/compartment

Conditions for “pre-marketing”

The definition of “pre-marketing” is augmented by the new conditions for pre-marketing. An EEA AIFM may engage in “pre-marketing”, except where the information provided to investors:

Is sufficient to allow investors to commit to acquiring interests in the AIF;
Amounts to subscription forms (or similar documents) whether in draft or final form;
Amounts to constitutional documents, a prospectus or offering documents of an AIF which has not yet been established, in final form.

In addition, any draft offering documents must not contain sufficient information to allow investors to take an investment decision and must include a statement that: (i) the document does not constitute an offer or an invitation to subscribe for interests in the AIF; and (ii) the information presented should not be relied upon because it is incomplete or subject to change. It is clear that simply labelling a document as draft but otherwise being in final form will not be acceptable. In addition, the circulation of subscription forms is prohibited and they must not be provided even in draft form during a “pre-marketing” exercise.

Notification of pre-marketing

It will be a requirement for EEA AIFMs to notify their NCA within two weeks of commencing “pre-marketing”. The notification will need to specify where and for which periods the pre-marketing is taking or has taken place with a brief description of the pre-marketing.

This notification of pre-marketing is distinct from the AIFMD marketing process and it will not be possible for an EEA AIFM to accept a subscription before the AIFMD marketing process has been completed. Importantly, any subscription made within 18-months of pre-marketing activity will be considered to be the result of marketing activity, triggering the AIFMD marketing process. Consequently, commencing pre-marketing will effectively preclude reliance by the AIFM on reverse solicitation for a period of 18-months.

Placement agents and other distributors

Significantly, any third party carrying out pre-marketing on behalf of an AIFM must be authorised in the EEA as an investment firm (or a tied agent of an investment firm), credit institution, UCITS management company or AIFM. This is an important change, particularly in light of Brexit. Placement agents and other distributors will need to consider whether they have the appropriate regulatory authorisations and seek to move EEA distribution into an appropriately authorised entity in advance of 2 August 2021.

Discontinuing marketing

There has been considerable uncertainty and diverge of practices across the EEA in relation to when an AIFM marketing under the AIFMD marketing passport can discontinue marketing. The CBD/R introduces a new formalised procedure for this. An AIFM will be able to discontinue marketing where certain conditions are fulfilled and a notification is made to the relevant NCA.

The conditions are:

Except in the case of closed-ended AIFs, making a blanket offer to repurchase or redeem interests held by investors in that Member State;
Publicising the intention to terminate marketing arrangements;
Terminating or modifying contracts with intermediaries/distributors to ensure they do not continue to market the AIF.

There is effectively a restriction on pre-marketing of successor AIFs once an AIF has been de-notified from marketing. The AIFM cannot engage in pre-marketing that AIF (or a successor AIF with similar investment strategy/idea) in that Member State for 36 months from the date of de-notification. Clearly this will cause difficulties, particularly for closed-ended funds. It is likely that AIFMs will simply choose not to de-register for marketing in order to not impact when the successor fund can be taken to market.

Impact on non-EEA sponsors

The CBD/R impacts EEA AIFMs directly. However, the changes will impact non-EEA managers in a range of ways:

EU Member States do not have a completely free hand as regards non-EEA AIFMs marketing in their jurisdiction under Article 42 of the AIFMD and their NPPRs. The CBD contains a recital that states that national rules cannot in any way disadvantage EEA AIFMs vis-à-vis non-EEA AIFMs. It is therefore highly likely that we will see a levelling-up of the NPPRs across the EEA with the requirements of the CBD/R on pre-marketing and discontinuation of marketing.
Non-EEA managers using an EEA host AIFM in order to avail themselves of the EEA marketing passport for a parallel vehicle will find the marketing of the fund subject to the new requirements. In particular, the requirement that any person marketing the fund on behalf of the EEA AIFM (i.e. the sponsor) must be authorised under the relevant sectoral legislation in the EU.

Application in the UK

The Financial Services (Implementation of Legislation) Bill 2017-19 provided a mechanism to implement EU financial services legislation that: (1) had been adopted by the EU, but did not yet apply; and (2) EU proposals that were in negotiation and that may have been adopted up to two years post-Brexit. The CBD/R were listed in the Schedule to the Bill. However, the Bill was not passed by the end of the 2017-19 parliamentary session and made no further progress. It is, therefore, currently uncertain how the CBD/R will be implemented in the UK post-Brexit.

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work in the firm’s Investment Funds, Private Equity or Financial Institutions practice groups, or the authors:

Michelle M. Kirschner – London (+44 (0)20 7071 4212, [email protected])
Martin Coombes – London (+44 (0)20 7071 4258, [email protected])

Please also feel free to contact any of the following practice leaders and members:

Investment Funds Group:
Michelle M. Kirschner – London (+44 20 7071 4212, [email protected])
Chézard F. Ameer – Dubai and London (+971 (0)4 318 4614, [email protected])
Albert S. Cho – Hong Kong (+852 2214 3811, [email protected])
John Fadely – Hong Kong (+852 2214 3810, [email protected])
Jennifer Bellah Maguire – Los Angeles (+1 213-229-7986, [email protected])
Y. Shukie Grossman – New York (+1 212-351-2369, [email protected])
Edward D. Sopher – New York (+1 212-351-3918, [email protected])
C. William Thomas, Jr. – Washington, D.C. (+1 202-887-3735, [email protected])
Gregory Merz – Washington, D.C. (+1 202-887-3637, [email protected])

Private Equity Group:
Scott Jalowayski – Hong Kong (+852 2214 3727, [email protected])
Sean P. Griffiths – New York (+1 212-351-3872, [email protected])
Ari Lanin – Los Angeles (+1 310-552-8581, [email protected])
Steven R. Shoemate – New York (+1 212-351-3879, [email protected])

Financial Institutions Group:
Arthur S. Long – New York (+1 212-351-2426, [email protected])
Matthew L. Biben – New York (+1 212-351-6300, [email protected])
Michael D. Bopp – Washington, D.C. (+1 202-955-8256, [email protected])
Stephanie Brooker – Washington, D.C. (+1 202-887-3502, [email protected])
M. Kendall Day – Washington, D.C. (+1 202-955-8220, [email protected])
Jeffrey L. Steiner – Washington, D.C. (+1 202-887-3632, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

The New York State Department of Financial Services (“DFS”) is the state’s primary financial regulating agency, with jurisdiction over approximately 1,500 financial institutions and 1,400 insurance companies. Moving into 2020, the agency is poised make its mark with a new administration that is focused on consumer protection in the absence of federal regulators, and on asserting its authority over emerging areas of significance to New York’s banking and insurance industries.

Just last month, the Governor proposed “sweeping” consumer protection measures as part of his 2020 agenda, and DFS announced the creation of a new Consumer Protection Task Force to help build support for and implement the Governor’s proposals. This month, DFS held consumer protection hearings to “hear directly from New York consumers about their experiences with financial products and services.” And the agency recently announced that it will hold a “Symposium on Financial Innovation and Inclusion” in April during New York Fintech Week, with the aim of bringing together various stakeholders to explore new developments and risks relating to emerging financial technology.

This DFS Round-Up summarizes recent key developments regarding the agency. Those developments are organized by subject.

To view the Round-Up, click here.

The following Gibson Dunn lawyers assisted in preparing this client update: Matthew Biben, Mylan Denerstein, Akiva Shapiro, Seth Rokosky, Bina Nayee and Zachary Piaker.

Gibson Dunn’s lawyers are available to assist with any questions you may have regarding these issues. For further information, please contact the Gibson Dunn lawyer with whom you usually work, any member of the firm’s Public Policy or Financial Institutions practice groups, or the following in New York:

Matthew L. Biben – Co-Chair, Financial Institutions Practice (+1 212-351-6300, [email protected])
Mylan L. Denerstein – Co-Chair, Public Policy Practice (+1 212-351-3850, [email protected])
Akiva Shapiro (+1 212-351-3830, [email protected])
Seth M. Rokosky (+1 212-351-6389, [email protected])

On 19 February 2020, the UK Supreme Court rendered its judgment in Micula and others v Romania.[1] In a unanimous ruling, the Supreme Court lifted a stay of enforcement of an ICSID arbitral award despite an extant State aid investigation by the European Commission (the Commission). With the stay lifted, the requirement for Romania to provide security as a condition of the stay was discharged.

This highly significant decision is the latest stage in the Micula brothers’ (the Claimants) longstanding attempts to enforce their ICSID Award against Romania, which has involved proceedings in numerous jurisdictions. The Court’s judgment concluded that while the English courts have the power to stay execution of ICSID awards in certain limited circumstances, the stay in this case exceeded the proper limits of that power. In particular, the EU Treaties did not displace the UK’s obligations under the ICSID Convention (pursuant to which the UK had a prior (pre-EU-accession) obligation to enforce the Award).

Background

In December 2013, an ICSID Tribunal issued an Award finding Romania in breach of the Sweden-Romania bilateral investment treaty,[2] and awarding the Claimants compensation of approximately £150 million (~$200 million), plus compound interest until satisfaction of the Award.[3] Romania subsequently commenced ICSID annulment proceedings, which were ultimately rejected.[4]

In October 2014, the Claimants applied for registration of the Award before the Commercial Court,[5] pursuant to the Arbitration (International Investment Disputes) Act 1966, which implemented the ICSID Convention into English law. In July 2015, Romania applied for the registration to be varied or set aside. By a counter-application, the Claimants sought an order for security to be made in the event that a stay of enforcement was ordered.

Meanwhile, in March 2015, the European Commission issued a decision (the Commission Decision) addressed to Romania concluding that payment of the Award constituted State aid in breach of the Treaty on the Functioning of the EU (the TFEU). Payment was, therefore, prohibited,[6] and any payments made hitherto were to be recovered. The Claimants applied to the General Court of the EU (the GCEU) for the Commission Decision to be annulled.

Returning to the UK enforcement efforts: the High Court refused to set aside registration, but granted a stay enforcement pending the outcome of the GCEU proceedings.[7] The High Court meanwhile refused the Claimants’ application for security.[8] In 2018, the Court of Appeal took a different approach. While maintaining the stay, it ordered Romania to provide security in the sum of £150 million.[9] Romania was permitted to appeal the order for security before the Supreme Court. The Claimants cross-appealed the grant of a stay.

In a further twist, on 18 June 2019—the morning the Supreme Court proceedings were scheduled to start—the GCEU annulled the Commission’s Decision (the Annulment Decision). The GCEU found that the Commission had exceeded its competence by retroactively applying its State aid powers under the TFEU to events predating Romania’s accession to the EU.[10] The GCEU’s judgment caused the Court of Appeal’s order of a stay (and the conditional security) to lapse. The Supreme Court hearing was postponed.

The Commission next made clear its intention to appeal the GCEU’s decision before the Court of Justice of the EU (CJEU) triggering further applications by Romania and the Claimants for the stay of enforcement and grant of security, respectively. The High Court ordered both.[11] A leapfrog application was granted for the parties’ appeals on these issues to be heard by the Supreme Court, with the Commission participating as an intervening party.[12]

The Supreme Court Judgment

The Supreme Court essentially considered three issues in relation to the stay, advanced by the Claimants. As the stay was lifted, it was unnecessary for the Court to consider Romania’s grounds of appeal against the order of security.

Did the Annulment Decision mean that the EU law duty of sincere cooperation no longer required the English courts to stay enforcement?

EU law provides for a “duty of sincere cooperation”,[13] which includes a mutual legal obligation for the EU and its Member States to assist each other in carrying out tasks which flow from the EU Treaties. It is intended to preserve the effectiveness of actions taken by EU bodies.

The issue for the Supreme Court was whether the Annulment Decision also annulled—as the Claimants argued—the Commission’s prior decision initiating formal State aid proceedings against Romania (on the basis it was “tainted by the same illegality”).[14] If so, no requirement existed for the stay to be maintained.

The Court disagreed. While the prior decision was subject to the “same flaws” as the Commission Decision, those flaws did not prevent the Commission from relying on it as giving rise to a duty of sincere cooperation on the part of the national courts. Moreover, the Annulment Decision left the State aid investigation open. Absent a final decision by the CJEU and a formal closure of the State aid investigation, the duty of sincere cooperation applies:[15] i.e., the English courts must ensure compliance with EU law.

Subject to the further grounds of appeal, the Court found the grant of stay should be upheld.

Do the English courts have the power to grant the stay, and is it incompatible with the ICSID Convention?

The UK signed the ICSID Convention in 1966, prior to its accession to the EU. Under the ICSID Convention, Article 54(1) imposes a duty on national courts to recognise an ICSID award as binding, and to enforce it “as if it were a final judgment” by a domestic court.

In the Claimants’ view, while national courts have control over the execution of an award (including the power to grant a temporary stay), it may only do so: (i) for procedural (not substantive) reasons; and (ii) where no inconsistency arises with the (Convention) duty to recognise and enforce the award. The English courts thus did not have the power to grant a stay pending determination of the GCEU proceedings.

The Court agreed. The Court first observed that “a notable feature” of the ICSID Convention was that “once the authenticity of an award is established, a domestic court … may not re-examine the award on its merits”; nor can a domestic court refuse to enforce on grounds of public policy.[16] Nevertheless, the Convention’s travaux préparatoires indicate that in “certain exceptional or extraordinary circumstances” national law defences to enforcement may be invoked.[17]

With that in mind, and taking into account the wording of Article 54(1), the Court concluded (agreeing with the majority in the Court of Appeal) that English courts have the power to stay execution in the “limited” circumstances described above.[18] In the present case, the Court found the stay “exceeded the proper limits of that power”, and “was not consistent with the ICSID Convention under which the [UK] and its courts had a duty to recognise and enforce the Award”.[19] The stay was not a limited stay on procedural grounds, but a prohibition on enforcement of the Award on substantive grounds.[20]

Do the UK’s EU law obligations require the UK to breach its pre-accession obligations under the ICSID Convention?

Article 351 TFEU provides that obligations arising from pre-EU-accession agreements involving third countries (i.e., non-EU Member States) “shall not be affected by the provisions of the EU Treaties”. In the Claimants’ view, the UK’s obligations to recognise and enforce awards under the ICSID Convention are pre-accession obligations within the meaning of Article 351. As such, they are unaffected by EU obligations and a stay is not required.

The Court agreed with the Claimants. While the UK, Sweden and Romania were at all times EU-Member States, under the ICSID Convention, obligations are owed not just to EU-Member States but to Contracting States which are third countries. Article 351 is, therefore, engaged. More specifically, the ICSID Convention confers specific duties—owed to all other Contracting States—to recognise and enforce awards (Article 54) and take such measures as may be necessary to implement the Convention (Article 69).[21] Nothing in the Convention or the travaux warranted viewing those specific duties as only owed to the State of nationality of an award beneficiary (which, in this case would mean only the rights of EU-Member States are engaged, and Article 351 would not apply).

In Romania’s (and the Commission’s) view, the Article 351 issue nevertheless required a stay in accordance with the duty of sincere cooperation. This is because of the risk of conflict with a future (CJEU) ruling.

The Court, however, disagreed; it was not required to defer to the EU courts on this issue. First, questions regarding prior treaties under Article 351 are not are not reserved to the EU courts.[22] Second, the Article 351 issue before the EU courts was different: it concerned Romania’s obligations, and not the UK’s obligations.[23] Third, the possibility that the EU courts may consider the issue at some future stage is “contingent and remote”.[24] Thus, the circumstances did not require a stay.

Accordingly, the Supreme Court allowed the Claimants’ appeal, and lifted the stay, which it described as “an unlawful measure in international law and unjustified and unlawful in domestic law”.[25] The issue of security consequently fell away.[26]

Comment

The Supreme Court’s judgment in Micula is significant. The Court agreed with the majority in the Court of Appeal that English courts have the power to stay execution of an ICSID award—but only in limited circumstances. Stays may only be granted for procedural (and not substantive) reasons, and only in circumstances where no inconsistency arises with the ICSID Convention duty on national courts to recognise and enforce the award. This is positive news for parties looking to enforce ICSID awards in the UK.

So far as Brexit is concerned, as matters currently stand, the UK remains bound by EU law. The relevance of the “duty of sincere cooperation” in future proceedings of a similar nature will depend on the agreement subsequently reached between the UK and the EU.

______________________

[1] Micula and others v Romania [2020] UKSC 5.

[2] Ioan Micula, Viorel Micula, S.C. European Food S.A, S.C. Starmill S.R.L. and S.C. Multipack S.R.L. v. Romania, ICSID Case No. ARB/05/20, Award, 11 December 2013.

[3] The Award plus interest is now estimated to be worth in excess of £251 million.

[4] Ioan Micula, Viorel Micula, S.C. European Food S.A, S.C. Starmill S.R.L. and S.C. Multipack S.R.L. v. Romania, ICSID Case No. ARB/05/20, Decision on Annulment, 26 February 2016.

[5] Micula and others v Romania [2020] UKSC 5, ¶ 28.

[6] Commission Decision 2015/1470 of 30 March 2015 on State aid SA.38517 (2014/C) (ex 2014/NN) implemented by Romania.

[7] Viorel Micula and others v Romania and European Commission (Intervener) [2017] EWHC 31 (Comm); [2017] Bus LR 1147.

[8] As per a subsequent judgment in Viorel Micula and others v Romania and European Commission (Intervener) [2017] EWHC 1430 (Comm).

[9] Viorel Micula and others v Romania and European Commission (Intervener) [2018] EWCA Civ 1801; [2019] Bus LR 1394.

[10] European Food SA and Others v European Commission (Cases T-624/15, T-694/15 and T-704/15) EU:T:2019:423.

[11] Viorel Micula and others v Romania and European Commission (Intervener) [2019] EWHC 2401 (Comm).

[12] Micula and others v Romania [2020] UKSC 5, ¶ 36.

[13] Micula and others v Romania [2020] UKSC 5, ¶ 42.

[14] Micula and others v Romania [2020] UKSC 5, ¶ 49. The Claimants further argued that the Annulment Decision annulled the effect of the Commission’s injunction decision of May 2014, which prohibited Romania from implementing the Award pending further investigation by the Commission. The Court concluded that it was not necessary to reach a conclusion on this issue. Micula and others v Romania [2020] UKSC 5, ¶ 52.

[15] Micula and others v Romania [2020] UKSC 5, ¶ 51.

[16] This may be contrasted with the New York Convention on the Recognition and Enforcement of Foreign Arbitral Awards 1958, which provides a detailed list of grounds on which recognition and enforcement

may be refused, including public policy grounds pursuant to Article V(2)(b). Micula and others v Romania [2020] UKSC 5, ¶ 68.

[17] Micula and others v Romania [2020] UKSC 5, ¶ 78.

[18] I.e., provided that in the particular circumstances it was just to do so and the stay was temporary and consistent with the purposes of the ICSID Convention. Micula and others v Romania [2020] UKSC 5, ¶ 84.

[19] Micula and others v Romania [2020] UKSC 5, ¶ 84.

[20] Micula and others v Romania [2020] UKSC 5, ¶ 84.

[21] ICSID Convention, Article 69 (“Each Contracting State shall take such legislative or other measures as may be necessary for making the provisions of this Convention effective in its territories.”).

[22] Micula and others v Romania [2020] UKSC 5, ¶ 112.

[23] Micula and others v Romania [2020] UKSC 5, ¶ 113.

[24] Micula and others v Romania [2020] UKSC 5, ¶¶ 114-117.

[25] Micula and others v Romania [2020] UKSC 5, ¶ 118.

[26] Micula and others v Romania [2020] UKSC 5, ¶ 119.

The following Gibson Dunn lawyers assisted in the preparation of this client update: Cyrus Benson, Jeffrey Sullivan, Piers Plumptre, Stephanie Collins and Theo Tyrrell.

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these issues. Please contact the Gibson Dunn lawyer with whom you usually work, any member of the firm’s International Arbitration Practice Group or any of the following:

International Arbitration Group:
Cyrus Benson – London (+44 (0) 20 7071 4239, [email protected])
Penny Madden – London (+44 (0) 20 7071 4226, [email protected])
Jeffrey Sullivan – London (+44 (0) 20 7071 4231, [email protected])
Rahim Moloo – New York (+1 212-351-2413, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

On February 13, 2020, final regulations went into effect to expand the scope of inbound foreign investment subject to review by the Committee on Foreign Investment in the United States (“CFIUS” or the “Committee”).

CFIUS is an inter-agency federal government group authorized to review the national security implications associated with foreign acquisitions of or investments in U.S. businesses and to block transactions or impose measures to mitigate any threats to U.S. national security. Until last year, the Committee’s jurisdiction was limited to transactions that could result in the control of a U.S. business by a foreign person. As we described here, the 2018 Foreign Investment Risk Review and Modernization Act (“FIRRMA”) expanded the scope of transactions subject to the Committee’s review to include certain non-controlling but non-passive foreign investments in U.S. businesses involved in critical technologies, critical infrastructure, or sensitive personal data of U.S. citizens (abbreviated as “TID” businesses for technology, investment, and data) as well as real estate transactions—including leases, sales, and concessions—involving air or maritime ports or in close proximity to sensitive U.S. government facilities.[1] Many of the critical issues set forth in FIRRMA were clarified by proposed regulations published by the U.S. Department of the Treasury on September 17, 2019, described here, and further refined in the final regulations published on January 13, 2020.

In a city that is rarely praised for efficiency or collaboration, the deliberative process that shaped these new rules merits some discussion. At the earliest stages of the legislative process, proposed CFIUS reform bills would have required scrutiny of innumerable transactions with no ostensible national security risk—including passive foreign investments through investment funds in ostensibly low risk industries and joint ventures with a foreign company partner. After months of negotiations, the House and Senate agreed upon legislation that expanded the scope of transactions subject to the Committee’s review, but punted key details to subsequent implementing regulations. Since proposing such regulations last year, the Committee sought and reviewed numerous written comments and requests for clarifications regarding the regulations in a transparent and public process. The final regulations reflect several changes made in response to such feedback, as well as lessons learned during the pilot program for mandatory filings involving certain types of critical technologies (the “Pilot Program”). The result is a smarter set of regulations designed to target real risks, as well as commentary that reflects the effort being made by the intelligence community to assess and adapt to increasingly complex investment structures.

Our top ten observations regarding these new regulations are set forth below.

1. Mandatory Filings for Critical Technology U.S. Business Transactions

The final regulations retain—with relatively minor changes—the Pilot Program’s mandatory fling requirement for certain transactions involving investments by foreign persons in U.S. businesses that deal in one or more “critical technologies.” The Pilot Program had served for the last year as a laboratory for the Committee to test out certain aspects of its newly expanded authorities—including mandatory pre-transaction filings for transactions involving critical technology U.S. businesses. These high-risk technologies include items subject to existing U.S. export controls, including emerging and foundational technologies to be identified pursuant to the Export Control Reform Act of 2018 (“ECRA”).

Under the new rules, transactions triggering mandatory CFIUS review will continue to include any investment by which a foreign person acquires material nonpublic technical information about the target critical technology U.S. business, membership or observer rights on the target’s board, or the right to participate in substantive decision-making, as well as transactions in which foreign persons acquire control of a critical technology U.S. business. Such businesses include those companies that produce, design, test, manufacture, fabricate, or develop certain items subject to the Export Administration Regulations (“EAR”), defense articles or defense services subject to the International Traffic in Arms Regulations (“ITAR”), “emerging and foundational technologies” that are to be identified through an interagency process chaired by the Department of Commerce going forward, as well as items subject to several other U.S. export control regimes.

To trigger the mandatory filing requirement under current regulations, the critical technology U.S. business in which the foreign person plans to invest must also operate in one of 27 high-risk industries identified by their five-digit North American Industry Classification System (“NAICS”) codes in Appendix B to Part 800. Notably, the Pilot Program illustrated that there is no definitive means for establishing the NAICS code applicable to a particular U.S. business, that a company’s “primary” NAICS code—which CFIUS often requests—may not capture the full scope of its business operations, and that companies also often have limited experience with evaluating the applicable NAICS codes or establishing a process for doing so. Rather than depend on this uncertain, unfamiliar metric for determining its jurisdiction, the Committee has indicated that it will eventually propose a new rule to replace the use of NAICS codes with a requirement based on export control licensing requirements. This change will likely make jurisdictional determinations more efficient and certain. The jurisdictional assessment for critical technology transactions already requires an evaluation of the target’s exposure to U.S. export controls. Additionally, determining export controls classifications and applicable license requirements is a common component of compliance for many companies dealing in critical technologies.

Forthcoming Commerce Department regulations to implement ECRA’s mandate will further clarify the range of companies that will be impacted by the Committee’s jurisdiction over critical technology business transactions. Under the Pilot Program and continuing under the new CFIUS regulations, “critical technologies” include items to be controlled as “emerging and foundational technologies” under new regulations the Commerce Department is required to promulgate. Observers have been expecting new regulations on emerging technologies—which will purportedly include new controls on particular kinds of artificial intelligence and quantum computing technology, among several other areas of emerging technology—to be published for months. (The Commerce Department has yet to publish a companion Advanced Notice of Proposed Rulemaking to solicit input on how to define and identify foundational technologies.) Commerce Department officials have repeatedly stated that publication of the new rules and controls on emerging technologies is imminent, but deliberations within the Trump administration appear to be delaying their publication. Depending on the schedule for publishing these rules and how the Commerce Department follows through on ECRA’s expressed preference for building international support to impose multilateral controls on specific technologies, it could be many months or even years before any specific definitions of emerging or foundational technologies are adopted.

For transactions subject to the CFIUS mandatory filing requirement, parties will continue to have the option of either filing a short-form declaration available on the Committee’s website or filing a full-length notice. In many cases, given the close scrutiny to which transactions involving critical technologies have recently been subject, a full-length notice may be advisable in order to reduce the total amount of time required for the Committee to complete its review, as described further below.

2. Mandatory Filings for Substantial Foreign Government Investments

In addition to the mandatory filing requirement for non-passive, non-controlling investments in a U.S. business dealing in critical technologies in connection with certain high-risk industries, filings are now also required for all transactions by which a foreign government obtains a “substantial interest” in a TID U.S. business. Specifically, a CFIUS filing is now required when a foreign government holds a 49 percent or greater voting interest in a foreign person that would obtain a 25 percent or greater voting interest in the target U.S. business. In the case of an entity with a general partner, managing member, or equivalent, the Committee will consider a foreign government to have a “substantial interest” if the foreign government holds 49 percent or more of the interest in the general partner, managing member, or equivalent. The new regulations further clarify that a “substantial interest” applies to a single foreign government, including both national and subnational governments, and their respective departments, agencies, and instrumentalities. In this regard, the Committee does not aggregate foreign governments’ interests when determining whether they are “substantial.”

3. New Exceptions to Mandatory Filings Requirements

In response to public comments, the final CFIUS regulations incorporate several new exceptions to mandatory filing requirements:

Foreign Ownership, Control, or Influence (“FOCI”) Mitigated Entities. The final CFIUS regulations incorporate an exception for investments by foreign investors operating under a valid facility security clearance and subject to an agreement to mitigate FOCI pursuant to the National Industrial Security Program regulations. Such FOCI mitigation agreements require the foreign entity holding a facility security clearance to implement an action plan to mitigate the security risk that the foreign ownership, control, or influence poses.
Exception for Investments Involving License Exception ENC. The regulations include a narrow exception for foreign investments in a U.S. business that would otherwise trigger a mandatory filing solely because the business produces, designs, tests, manufactures, fabricates, or develops one or more critical technologies that are eligible for License Exception ENC under the EAR, which authorizes the export of encryption commodities, software, and technology without a specific government-issued license. Helpfully, this exception will apply to many software and technology companies that include different kinds of encryption functionality in their products.

As described further below, the final CFIUS regulations retain exceptions from the mandatory filing requirement for passive and indirect foreign investment made through investment funds, as well as covered investments by certain “excepted” investors from exempted foreign states.

4. Declarations for All Transactions—Still No Filing Fees

In order to help CFIUS and the regulated public manage the burden of CFIUS’s expanded remit, FIRRMA provided a new-short form filing—the “declaration.” These declarations are built from a standard five-page form, available on CFIUS’s website. They require similar, but less extensive, information about the proposed transaction than the standard notice and are subject to an abbreviated 30-day review period.

Over the last year, CFIUS has test-driven this new filing tool in its Pilot Program. Parties to a Pilot Program covered transaction—including covered investments in and acquisitions of critical technology U.S. businesses—were required to file with the Committee in advance of closing the transaction and could choose to submit either a short-form declaration or the traditional long-format notice. But the Pilot Program proved a poor testing ground for the new tool. Perhaps because of the national security concerns inherent in Pilot Program covered transactions—which dealt exclusively with companies handling export controlled technology for sensitive industries—or because of the Committee’s expanded case load, CFIUS was frequently unable to clear declaration cases within the shortened 30-day review period. As a result, CFIUS would often request parties subsequently file a notice—increasing the amount of work required of the parties and dramatically extending the total CFIUS review timeline. By some estimates, only 10 percent of cases filed with CFIUS using the new short-form declaration were cleared in the shortened review period.

Now the declaration is getting a wider release. Under the new regulations, parties to any transaction subject to CFIUS review are permitted to submit a short-form declaration as an alternative to the lengthier voluntary notice procedure that is subject to an expedited review process. The declaration may become more useful as parties to relatively low-risk transactions can now opt for the short-form filing and shorter review timeline. The Committee itself has cautioned that parties should consider the likelihood that CFIUS will be able to conclude action in the 30 days allotted for reviewing a declaration when determining which format to file, suggesting that the long-form notice may be more appropriate for more complex transactions bearing indicia of national security risk. Interestingly, neither submission format is yet subject to a filing fee, although fees will likely be proposed pursuant to forthcoming regulations.

5. Preliminary List of Excepted Foreign States

The new CFIUS regulations create an exception from certain real estate transactions and non-controlling TID investments (but not transactions that could result in control) for investors based on their ties to certain countries identified as “excepted foreign states,” and their compliance with certain laws, orders, and regulations (including U.S. sanctions and export controls).[2] Although CFIUS initially suggested that it would be several years before such excepted foreign states were named, the Committee’s final regulations defied observers’ expectations by exempting investors from three named countries—Australia, Canada, and the United Kingdom—from CFIUS scrutiny in certain circumstances. Although the timing of the announcement surprised observers, the countries it selected were of no surprise. Australia, Canada, and the United Kingdom are all members of the multilateral UKUSA Agreement under which the so-called “Five Eyes”—the United States, Australia, Canada, New Zealand, and the United Kingdom—share intelligence data. According to the Committee these countries were selected due to aspects of their robust intelligence-sharing and defense industrial base integration mechanisms with the United States.

A country’s status as an excepted foreign state is conditioned upon the implementation of their own process to analyze foreign investments for national security risks and to facilitate coordination with the United States on matters relating to investment security, by February 13, 2022. This two-year grace period will provide the Committee with time to develop processes and procedures to determine whether the standard has been met with respect to other jurisdictions. Although the Committee, indicated that it takes no current position as to whether these foreign states currently meet the review process requirement, all three named countries either have or are formalizing investment review processes.

Although CFIUS could expand the list of “excepted foreign states” in advance of 2022, few other allies are as closely interwoven with the United States as Australia, Canada or the United Kingdom. Moreover, the new regulations eliminate language in the proposed rules that would have allowed an eligible country to be selected with a consensus of two-thirds of CFIUS’s voting members. As a result, selection of an additional eligible country likely would require consensus among the entire Committee. Furthermore, the final regulations warn that an expansive application carries potentially significant implications for the national security of the United States, suggesting that a “go-slow” approach to expanding the list of excepted foreign states is likely to prevail.

6. Clarifying the Excepted Investor Provision

The excepted investor provision was designed to accommodate increasingly complex ownership structures in the application of the Committee’s jurisdiction, and generally exempts persons, governments, and entities from excepted foreign states from certain types of CFIUS scrutiny. As mandated by FIRRMA, this exception was intended to limit the expansion of CFIUS jurisdiction. Notably, the final regulations revised the earlier proposed definition of excepted investor in response to numerous comments which suggested relaxing the criteria with respect to the nationality of board members and observers, the percentage ownership limits for individual investors, and the minimum excepted ownership.

The final regulations indicate that a foreign entity will qualify as an excepted investor if it meets each of the following conditions with respect to itself and each of its parents: (i) such entity must be organized under the laws of an excepted foreign state or the United States, (ii) such entity must have its principal place of business in an excepted foreign state or the United States, (iii) 75 percent or more of the members and 75 percent or more of the observers of the board of directors or the equivalent governing body of such entity are U.S. nationals or nationals of one or more excepted foreign states who are not also nationals of any non-excepted foreign state, (iv) any foreign person that individually, and each foreign person that is part of a group of foreign persons that in the aggregate holds 10 percent or more of the outstanding voting interest of such entity (or otherwise could control such entity) is a foreign national of one or more excepted foreign states who is not also a national of any non-excepted foreign state; a foreign government of an excepted foreign state; a foreign entity that is organized under the laws of an excepted foreign state and has its principal place of business in an excepted foreign state or in the United States; and (v) the minimum excepted ownership of such entity is held, individually or in the aggregate, by one or more persons each of whom is (A) not a foreign person; (B) a foreign national who is a national of one or more excepted foreign states and is not also a national of any foreign state that is not an excepted foreign state; (C) a foreign government of an excepted foreign state; or (D) a foreign entity that is organized under the laws of an excepted foreign state and has its principal place of business in an excepted foreign state or in the United States.

The final rules increased the number of foreign nationals that may be on an excepted company’s board, raised the percentage ownership limit for individual investors in an excepted investor entity, and allow investors to have more foreign ownership than under the earlier proposed rules and still qualify for excepted status. However, the exception remains significantly cabined by the multiple layers of criteria it includes, and it remains possible for investors to lose excepted investor status and for their investments to become subject to CFIUS review. Additionally, CFIUS rejected commenters’ requests for a separate exception for “repeat customers” of the Committee who have previously or routinely obtained clearance and remain in good stead.

Notably, an investor’s nationality is not dispositive—the regulations identify additional criteria that a foreign person must meet in order to qualify for excepted investor status. Among these, investors cannot qualify for and may lose their excepted status if they are parties to settlement agreements with the U.S. Department of the Treasury’s Office of Foreign Assets Control (“OFAC”) or the U.S. Department of Commerce Bureau of Industry and Security (“BIS”), or are debarred by the U.S. Department of State, for sanctions or export control violations.

7. Investment Fund Carve-Out

The new rules also place limitations on the Committee’s jurisdiction with respect to passive and indirect foreign investments made through investment funds in TID U.S. businesses. An indirect investment by a foreign person in a TID U.S. business through an investment fund that affords the foreign person membership as a limited partner on an advisory board of the fund will not be considered a covered investment if certain conditions are met. First, the fund must be managed exclusively by a general partner or equivalent that is not a foreign person. Further, the advisory board membership must not afford the foreign person the ability to control the fund, participate in substantive decision-making regarding the fund, or access material nonpublic technical information. In its discussion of the new regulations, CFIUS also makes clear that this exception is limited and is not intended to create a presumption that any indirect investment by a foreign person in a TID U.S. business through an investment fund is a covered transaction if these criteria are not met. Instead, TID business investments will need to be analyzed on a case-by-case basis.

In the preamble to the new regulations, CFIUS indicated that numerous commenters requested further clarification regarding the scope of the investment carve-out, recommending revisions to the definition of “foreign entity” to focus on control by foreign persons, requesting additional examples of the types of limited partner rights that would not give rise to control. Citing the limitations of its authority under FIRRMA, the Committee declined to make most of these suggestions, opting instead to set forth a new interim rule defining “principal place of business” for CFIUS purposes, a measure that will address investment funds managed and controlled by U.S. persons in the United States, among other issues. The Committee did not adopt suggestions to apply the minimum excepted ownership criteria only to the general partner in a fund setting, noting that investment fund structures can vary significantly and limited partners may have significant rights vis-à-vis their investment interests.

As described further below, the new regulations provide an exemption to the mandatory filing requirement for investment funds controlled and managed by U.S. nationals. The regulations clarify, however, that a limited partner in a fund could have a filing obligation separate and apart from that of the fund. If a limited partner, for example, is granted control rights or access to material nonpublic technical information of a TID U.S. business, the limited partner may be subject to its own mandatory filing even if the fund itself is not.

8. New Principal Place of Business Interim Rule

In response to public comments seeking greater clarity about which entities are subject to CFIUS jurisdiction (including the aforementioned investment fund carve-out), the Treasury Department in January 2020 issued an interim rule defining an entity’s principal place of business as “the primary location where an entity’s management directs, controls, or coordinates the entity’s activities, or, in the case of an investment fund, where the fund’s activities and investments are primarily directed, controlled, or coordinated by or on behalf of the general partner, managing member, or equivalent.” Until now, the term principal place of business—which bears on whether an entity is “foreign” and thus subject to CFIUS jurisdiction—was undefined. The new definition, proposed in January 2020, was subject to a 30-day public comment period that ended on February 18, 2020.

That definition broadly tracks the test used by U.S. federal courts for determining diversity jurisdiction, in which the court looks to where the corporate “nerve center” is located. The definition of principal place of business also includes a special rule designed to ensure “consistent treatment of an entity’s principal place of business in accordance with its own assertions to government entities, provided the facts have not changed since those assertions.” The new CFIUS definition of principal place of business therefore contains a second prong which provides that if an entity has represented to a U.S. federal, state, local or foreign government in its most recent submission or filing with that authority that its principal place of business is outside the United States, then that location will be deemed the entity’s principal place of business unless the entity can show that such location has since changed to the United States. From a policy standpoint, this carve-out appears designed to prevent entities from having their cake and eating it too—for example, by claiming to be based overseas for tax purposes, while also claiming to be U.S.-based for CFIUS purposes. This is an important clarification for foreign incorporated companies traded on U.S. exchanges, as the Committee continues to assert that U.S. shareholder addresses do not necessarily demonstrate that the owners of stock are U.S. nationals.

The final FIRRMA regulations were modified in several significant ways in response to comments that were received during the last comment period, and we expect a similarly robust dialogue between the Committee and the business community regarding this new proposed definition.

9. Personal Data Collections

The new rules also extended CFIUS jurisdiction to include review of certain investments in U.S. businesses that maintain or collect certain categories and quantities of sensitive data that may be exploited in a manner that threatens national security. This new category of jurisdiction covers U.S. businesses that (i) target or tailor products or services to sensitive populations, including U.S. military members and employees of federal agencies involved in national security; (ii) collect or maintain sensitive personal data on at least one million individuals; or (iii) have a “demonstrated business objective” to maintain or collect such data on greater than one million individuals with such data representing an integrated part of a U.S. business’s primary products or services. “Sensitive personal data” can include, among other types, financial data, geolocation data, U.S. government personnel clearance data, or biometric information that is maintained or collected by U.S. businesses described in (i)-(iii), above. Information derived from the results of genetic testing is considered “sensitive personal data” regardless of whether the business holding it is also described in (i)-(iii), above. Investments that provide a foreign person with certain information or governance rights with respect to such sensitive data U.S. businesses now trigger CFIUS jurisdiction.

In its final rules, the Treasury Department made few changes to this expanded jurisdiction in response to comments it received. Some commenters had expressed concern that the scope of information CFIUS considered to be sensitive personal data was too broad and would exceed what is necessary to protect national security. CFIUS refined the rule to clarify that data collected by U.S. businesses on their own employees, with certain exceptions for federal employees and contractors, and data that is a matter of public record does not quality for CFIUS review. A further clarification was made to narrow the scope of financial data covered by CFIUS to include only data that could be used to determine an individual’s financial distress or hardship.

There is still significant uncertainty regarding CFIUS’s potential use of this new rule, and it will likely be some time before its scope is fully understood. However, given the central and often commonplace role that data collection plays in many of today’s businesses, its application could be quite broad. What is clear from the promulgation of these new regulations is that any company with even moderate data collection practices will have to consider the potential impact of CFIUS on covered transactions. The threshold of collection or maintenance of data on at least one million individuals, or the demonstrated business objective to do so, is unlikely to provide a meaningful barrier to CFIUS review in many situations.

There are some initial indications that CFIUS will broadly interpret its jurisdiction over transactions involving sensitive personal data in a way that could affect many companies that may be unaccustomed to the challenges of navigating U.S. trade controls. For example, in March 2019, the Committee ordered Beijing Kunlun Tech Co. Ltd. (“Kunlun”) to sell its interest in Grindr LLC, a popular dating application focused on the LGBTQ community. Kunlun, a Chinese technology firm, acquired an approximately 60 percent interest in Grindr in January 2015 and subsequently completed a full buyout of the company in January 2018. Although CFIUS did not comment publicly, observers have speculated that the action was prompted over the Chinese firm’s access to sensitive personal data from Grindr users—such as location, sexual preferences, HIV status, and messages exchanged via the app. The Committee similarly intervened with Shenzhen-based iCarbonX after it acquired a majority stake in PatientsLikeMe, an online service that helps patients find people with similar health conditions within a user-base of around 700,000 people. As these cases suggest, CFIUS’s new jurisdiction over sensitive data U.S. businesses could affect a wide range of technology companies and service providers that have not typically been subject to CFIUS review or particularly burdensome U.S. export controls and may not be adequately prepared for such scrutiny.

10. Real Estate Transactions

As expected, the newly promulgated rules also expand CFIUS jurisdiction into new territory, including real estate transactions. Specifically, the Committee now has jurisdiction over the purchase or lease by, or concessions to, a foreign person of U.S. real estate that is within a defined range of certain airports, maritime ports, U.S. military installations, and other sensitive government sites listed in Appendix A to Part 802.

In a departure from CFIUS jurisdiction under prior rules, the Committee has jurisdiction to review real estate transactions even when the transaction does not involve a “U.S. business.” Importantly, however, even if the real estate at issue is not covered under the Committee’s expanded jurisdiction or if the foreign person did not acquire sufficient property rights as described below, CFIUS could still have jurisdiction over the transaction if it involves the transfer of control over (or a qualifying investment in) a U.S. business.

Covered real estate transactions include property associated with maritime ports and major domestic airports, and property located within “close proximity” or the “extended range” of certain military installations and other sensitive government sites identified in an appendix to the regulations. “Close proximity” is defined as property within one mile of the boundary of such facilities; whereas “extended range” is generally defined as property within 99 miles of identified government locations. CFIUS anticipates making a web-based tool available to help the public understand the geographic coverage of the new rule.

Additional limitations narrow the scope of what real estate transactions are covered. For example, a foreign person must acquire specified property rights in “covered real estate” to trigger CFIUS scrutiny—namely, the foreign person must be afforded as a result of the transaction three or more of the following property rights: (i) to physically access; (ii) to exclude; (iii) to improve or develop; or (iv) to affix structures or objects. The regulations leave undefined lesser rights. Importantly, the right or ability to determine the type of development to occur on the property, or to participate in decisions regarding tenants or leases, or to monitor the property likely would not trigger CFIUS jurisdiction.

Furthermore, specific exceptions apply to certain properties in urban areas. Real estate in “urbanized areas” and “urban clusters” as defined by the Census Bureau in the most recent U.S. census are excluded from the Committee’s real estate jurisdiction. Urban clusters are those territories with between 2,500 and 50,000 individuals whereas urbanized areas are those with more than 50,000 people. The urbanized area exclusion applies to covered real estate everywhere except where it is in “close proximity” to a military installation or sensitive U.S. government facility or where it will function as part of an airport or maritime port. Real estate transactions regarding single housing units and certain commercial office space are also excepted under the new rules.

Conclusion

If past is prologue, the Committee’s enforcement efforts in the coming months will highlight the types of risks that these new regulations are designed to target. In 2019, CFIUS forced several foreign companies to divest from U.S. businesses involved in the collection of sensitive personal data or cybersecurity, two issues likely to remain in the Committee’s crosshairs. Buoyed by new funding and personnel, we expect the Committee to proactively monitor the market for similarly high-risk transactions in the coming year.

_________________________

[1] FIRRMA was incorporated into the John S. McCain National Defense Authorization Act for Fiscal Year 2019, which was signed into law by President Trump on August 13, 2018.

[2] §800.219 (excepted foreign state); §800.220 (excepted investor); §802.215 (excepted real estate foreign state); §802.216 (excepted real estate investor).

The following Gibson Dunn lawyers assisted in preparing this client update: Judith Alison Lee, Jose Fernandez, Stephanie Connor, Chris Timura, R.L. Pratt, Scott Toussaint, Allison Lewis, JeanAnn Tabbaa, Brian Williamson, Cate Harding and Tory Roberts.

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding the above developments. Please contact the Gibson Dunn lawyer with whom you usually work, the authors, or any of the following leaders and members of the firm’s International Trade practice group:

United States:
Judith Alison Lee – Co-Chair, International Trade Practice, Washington, D.C. (+1 202-887-3591, [email protected])
Ronald Kirk – Co-Chair, International Trade Practice, Dallas (+1 214-698-3295, [email protected])
Jose W. Fernandez – New York (+1 212-351-2376, [email protected])
Marcellus A. McRae – Los Angeles (+1 213-229-7675, [email protected])
Adam M. Smith – Washington, D.C. (+1 202-887-3547, [email protected])
Stephanie L. Connor – Washington, D.C. (+1 202-955-8586, [email protected])
Christopher T. Timura – Washington, D.C. (+1 202-887-3690, [email protected])
Ben K. Belair – Washington, D.C. (+1 202-887-3743, [email protected])
Courtney M. Brown – Washington, D.C. (+1 202-955-8685, [email protected])
Laura R. Cole – Washington, D.C. (+1 202-887-3787, [email protected])
R.L. Pratt – Washington, D.C. (+1 202-887-3785, [email protected])
Samantha Sewall – Washington, D.C. (+1 202-887-3509, [email protected])
Audi K. Syarief – Washington, D.C. (+1 202-955-8266, [email protected])
Scott R. Toussaint – Washington, D.C. (+1 202-887-3588, [email protected])

Europe:
Peter Alexiadis – Brussels (+32 2 554 72 00, [email protected])
Nicolas Autet – Paris (+33 1 56 43 13 00, [email protected])
Attila Borsos – Brussels (+32 2 554 72 10, [email protected])
Patrick Doris – London (+44 (0)207 071 4276, [email protected])
Sacha Harber-Kelly – London (+44 20 7071 4205, [email protected])
Penny Madden – London (+44 (0)20 7071 4226, [email protected])
Steve Melrose – London (+44 (0)20 7071 4219, [email protected])
Benno Schwarz – Munich (+49 89 189 33 110, [email protected])
Michael Walther – Munich (+49 89 189 33-180, [email protected])
Richard W. Roeder – Munich (+49 89 189 33-160, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

On February 7, 2020, California Attorney General Xavier Becerra released a revised set of proposed regulations for the California Consumer Privacy Act of 2018 (“CCPA”), and released an additional amendment on February 10, 2020.[1] These proposed regulations provide further details and clarifications on the steps businesses must take to comply with the CCPA. This is not the end of the road for the development of the regulations, however, as the Attorney General will at least consider additional comments, which must be submitted by February 25, 2020, before the regulations are finalized.[2]

The CCPA[3] took effect January 1, 2020, and aims to give California consumers increased visibility into and control over how companies use and share their personal information. It applies to all entities doing business in California and collecting California consumers’ personal information if they meet certain thresholds. The Attorney General’s power to enforce the law is delayed until July 1, 2020. More information can be found in our prior client alerts on the topic, including a summary of the statute (here), amendments from October 2018 (here), additional proposed amendments (here), the Attorney General’s draft regulations (here), the final amendments signed in October 2019 (here), and a summary of CCPA developments heading into 2020 (here).

The revised version of the proposed regulations adjusts some of the requirements imposed on businesses by the initial proposed regulations, clarifies certain definitional ambiguities, and includes additional proposed provisions relating to service providers’ handling of personal information.

Below, we briefly summarize a number of the key changes in the revised proposed regulations. The list is not exhaustive, and we encourage you to contact us with any questions. As the public comment period is an important opportunity for companies to provide feedback to shape the proposed regulations, please feel free to contact any of the Gibson Dunn attorneys listed below if you are interested in submitting comments in advance of the February 25, 2020 deadline.

Key Definitions Clarified

“Personal information”: Version 2.0 of the proposed regulations (“Version 2.0”) adds guidance for interpreting the definition of “personal information” under the CCPA, alleviating some concern regarding exactly how broadly “personal information” would be applied. Specifically, whether information constitutes “personal information” depends on “whether the business maintains [the] information in a manner that identifies, relates to, describes, is reasonably capable of being associated with, or could reasonably be linked, directly or indirectly, with a particular consumer or household.”[4] The revised regulations clarify that IP addresses—which have been a particular focus for companies collecting statistical and analytical information on the usage of their websites—that are not tied to any identifiable consumers or households, and that cannot be reasonably linked to any identifiable consumer or household, do not constitute “personal information” under the CCPA in those instances.[5]
“Categories”: Version 2.0 clarifies that businesses must describe “categories of sources”[6] and “categories of third parties”[7] to consumers in notices at collection, privacy policies, and in response to verified requests to know with “enough particularity to provide consumers with a meaningful understanding of the type” of source or third party.

Notices At Collection Must Be Readily Accessible

The revised regulations update obligations related to notices at collection, particularly with respect to mobile applications. For instance, the regulations state notices must be posted wherever personal information is collected, including on webpages, mobile application download pages (and within mobile applications, such as the settings menu), and printed forms.[8] Furthermore, businesses must provide consumers with a “just-in-time” notice, describing the categories of personal information being collected and a link to the full notice at collection, when collecting personal information that consumers would “not reasonably expect” to be collected from mobile devices.[9] If personal information is collected orally, oral notice may be given.[10] In addition, the draft provides the following regarding notices:

No Additional Consent Required For Use Not “Materially” Different

Version 2.0 makes clear that additional consent is not required for the use of previously collected personal information that is not “materially” different from the uses disclosed in the original notice at collection. Under the previous iteration of the proposed regulations, any additional use of personal information that did not fall strictly into the uses described in the notice at collection would have required the business to seek additional consent.[11]

Specific Business Or Commercial Purpose Need Not Be Explicitly Tied To The Category Of Personal Information

The revised regulations no longer require a business to identify the specific business or commercial purpose for the collection of each category of personal information collected.[12] In other words, Version 2.0 suggests it is sufficient simply to list the business or commercial purposes for collecting all the categories of personal information collected, and a breakdown by category is no longer necessary. This revision is similarly explained in the section of Version 2.0 on Privacy Policies.[13]

Data Broker Obligations Simplified

Data brokers registered with the Attorney General under California’s data broker registration law (Civil Code § 1798.99.80) that post links to their privacy policies containing instructions on how to opt-out need not provide notices at collection.[14] This provision replaces the previously proposed mandates requiring businesses not collecting personal information directly from consumers to either (1) contact consumers directly to provide notice, or (2) contact the source of the information for an attestation describing how the source provided the required notice at collection (this provision was discussed in more depth in our previous client alert regarding the first draft of the regulations, available here). However, the new provision leaves unclear how “data scrapers” that do not sell personal information—or simply companies that obtain non-exempted personal information from sources other than the consumer, such as publicly available sources other than government records—should provide notice to the consumer.

Employee Notice Explained

The notice at collection provided to employees does not need to include a “Do Not Sell My Personal Information” link, and may include a link to (or a paper copy of) the employee privacy policy, instead of the general consumer privacy policy.[15]

“Do Not Sell” Provisions And Optional Opt-Out Button

Version 2.0 provides an option to obtain user consent to sell data that the business collected during the time it did not have a notice of the right to opt-out, as opposed to the total ban of that sale under the previous version of the proposed regulations.[16] The new draft regulations also give the option of providing an “opt-out” button alongside the “Do Not Sell My Personal Information” link, and provide a visual depiction of how such a button may appear (see below), but requires the link nonetheless.[17]

Responding To Requests To Know And Requests To Delete

Required Response Time Revised

Under the new proposed regulations, businesses are required to confirm receipt of requests to know and requests to delete within 10 business days—instead of “10 days”—though the allowed period to respond to the requests remains 45 calendar days.[18]

Designated Methods For Submitting Requests Simplified

The updated proposed regulations eliminate the requirement to maintain a webform. Businesses that operate solely through a website must provide an email address to submit requests to know, but no longer need to additionally maintain a webform. All other businesses must provide at least two designated methods for submitting requests to know, including at a minimum a toll-free number.[19] The requirement to provide two designated methods for submitting requests to delete (which do not necessarily include a webform), remains unchanged.[20]However, the proposed regulations indicate that businesses should still consider the ways in which they primarily interact with consumers when providing additional methods for submitting requests.

Exemption For Businesses That Do Not Sell And Only Maintain Personal Information For Legal Compliance

In response to consumers who make requests to know the personal information a business has collected about them, Version 2.0 relieves businesses of the requirement to search for personal information if the following conditions are met: 1) the business does not maintain personal information in a searchable or reasonably accessible format; 2) the business maintains the personal information solely for legal or compliance purposes; 3) the business does not sell personal information and does not use it for any commercial purpose; and 4) the business describes to the consumer the categories of records that may contain personal information, despite not having searched because it meets the above conditions.[21]

Biometric Information Should Not Be Provided In Response To Requests To Know

In response to requests to know, biometric information joins other categories of sensitive information that businesses cannot disclose, such as Social Security numbers and financial account numbers. Biometric data includes such information either generated from measurements or technical analysis of human characteristics. This is consistent with the legislature’s recent revision of the categories of information that trigger breach notice requirements, and consequently, the relevant categories of personal information subject to the private right of action under the CCPA.

Unverified Requests To Delete Do Not Need To Be Treated As Opt-Out Of Sales

Unlike the previous draft of the proposed regulations, businesses no longer need to treat unverified requests to delete as opt-out of sales of personal information. Instead, businesses are permitted to ask consumers if they would instead like to opt-out of the sale of their personal information, provided they had not already made a request to opt-out.[22]

Businesses Can Retain Record Of Requests To Delete

Version 2.0 explicitly allows businesses to retain a record of the request for the purpose of ensuring the consumer’s personal information remains deleted from the business’s records.[23]

Verification For Requests from Households With MinorsIf a member of the household is a minor under the age of 13, the business must obtain verifiable parental consent before complying with requests to access or delete specific pieces of information for the household.[24]
Verification For Non-Account Holders

Version 2.0 provides additional examples of acceptable methods for verifying consumers who do not have password-protected accounts, including asking the consumer to respond to in-app questions, or provide additional information about a transaction amount or an item purchased.[25]

Authorized Agent Requests

The new draft regulations allow businesses to require agents to submit a signed permission and consumers to directly confirm the authorization of the agent with the business, and impose additional requirements on authorized agents, such as implementing and maintaining reasonable security to protect consumers and restricting their use of the information.[26]

Service Providers Are Granted Greater Leeway

Version 2.0 no longer bars service providers from using the personal information they collect for their own purposes, so long as the personal information is not used to build household or consumer profiles or “clean or augment the data with data acquired from another source.”[27]

Furthermore, in response to requests to know or delete, service providers can either act on behalf of the business or inform the consumer that the request cannot be processed because they are a service provider.[28]

Clarifications Regarding Requests To Opt-Out

The updated regulations now allow businesses to propose an opt-in, after consumers have already opted-out of the sale of their personal information, if those consumers have attempted to use a product or service that requires the sale of their personal information.[29]

Recordkeeping Requirements Lessened

Version 2.0 changes the threshold triggering the recordkeeping requirement for businesses that collect, use, or disclose the personal information of large numbers of consumers within one year.[30] The threshold for triggering the recordkeeping requirement was increased from the collection of information from 4 million consumers to 10 million consumers. Businesses that meet the 10 million threshold must compile and disclose within their privacy policy the numbers for all requests to know, delete, and opt-out received for all individuals. Though the initial draft regulations required those businesses to report the number of requests received from California residents (consumers) specifically, Version 2.0 grants the option of disclosing the number of requests received from all individuals, eliminating the added effort that may be required to parse out California residents.[31]

Conclusion

Version 2.0 of the proposed CCPA regulations provides some much needed clarification on certain of the ambiguities in the CCPA. However, not all ambiguities have been resolved. For example, the new draft regulations do not provide any practical clarity on the prohibition of the use of personal information for the purpose of building a “consumer” or “household profile” by service providers.

Companies subject to the CCPA should continue to monitor the proposed regulations as they evolve. It is also important to provide comments and weigh in by February 25, 2020 on issues of interest to particular companies that remain unclear. We are available to assist with your inquiries as needed.

_____________________

[1] The entire text of the draft regulations is available at https://www.oag.ca.gov/sites/all/files/agweb/pdfs/privacy/ccpa-text-of-mod-redline-020720.pdf?.

[2] Department of Justice, Title 11, Division 1, Chapter 20. California Consumer Privacy Act Regulations (February 10, 2020), available at https://oag.ca.gov/sites/all/files/agweb/pdfs/privacy/ccpa-notice-of-mod-020720.pdf.

[3] The CCPA is encoded in California Civil Code Sections 1798.100 to 1798.198.

[4] Draft Regulations § 999.302(a) (emphasis added).

[5] Id.

[6] Draft Regulations § 999.301(d).

[7] Draft Regulations § 999.301(e).

[8] Draft Regulations § 999.305(a)(3).

[9] Draft Regulations § 999.305(a)(4).

[10] Draft Regulations § 999.305(a)(3)(d).

[11] Draft Regulations § 999.305(a)(5).

[12] Draft Regulations § 999.305(b)(2).

[13] Draft Regulations § 999.308(c)(1)(d).

[14] Draft Regulations § 999.305(d).

[15] Draft Regulations § 999.305(e).

[16] Draft Regulations § 999.306(e).

[17] Draft Regulations § 999.306(f).

[18] Draft Regulations § 999.313(a).

[19] Draft Regulations § 999.312(a).

[20] Draft Regulations § 999.312(b).

[21] Draft Regulations § 999.313(c)(3).

[22] Draft Regulations § 999.313(d)(1).

[23] Draft Regulations § 999.313(d)(3).

[24] Draft Regulations § 999.318(c).

[25] Draft Regulations § 999.325(e).

[26] Draft Regulations § 999.326(a), (e).

[27] Draft Regulations § 999.314(c). Please note that “household or consumer profiles” are not defined.

[28] Draft Regulations § 999.314(e).

[29] Draft Regulations § 999.316(b).

[30] Draft Regulations § 999.317(g).

[31] Id.

The following Gibson Dunn lawyers assisted in the preparation of this client update: Alexander Southwell, Ryan Bergsieker, Cassandra Gaedt-Sheckter, Abbey Barrera, and Lisa Zivkovic.

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work, or any member of the firm’s California Consumer Privacy Act Task Force or its Privacy, Cybersecurity and Consumer Protection practice group:

California Consumer Privacy Act Task Force:
Ryan T. Bergsieker – Denver (+1 303-298-5774, [email protected])
Cassandra L. Gaedt-Sheckter – Palo Alto (+1 650-849-5203, [email protected])
Joshua A. Jessen – Orange County/Palo Alto (+1 949-451-4114/+1 650-849-5375, [email protected])
H. Mark Lyon – Palo Alto (+1 650-849-5307, [email protected])
Alexander H. Southwell – New York (+1 212-351-3981, [email protected])
Deborah L. Stein (+1 213-229-7164, [email protected])
Eric D. Vandevelde – Los Angeles (+1 213-229-7186, [email protected])
Benjamin B. Wagner – Palo Alto (+1 650-849-5395, [email protected])

Please also feel free to contact any member of the Privacy, Cybersecurity and Consumer Protection practice group:

United States
Alexander H. Southwell – Co-Chair, PCCP Practice, New York (+1 212-351-3981, [email protected])
Debra Wong Yang – Los Angeles (+1 213-229-7472, [email protected])
Matthew Benjamin – New York (+1 212-351-4079, [email protected])
Ryan T. Bergsieker – Denver (+1 303-298-5774, [email protected])
Howard S. Hogan – Washington, D.C. (+1 202-887-3640, [email protected])
Joshua A. Jessen – Orange County/Palo Alto (+1 949-451-4114/+1 650-849-5375, [email protected])
Kristin A. Linsley – San Francisco (+1 415-393-8395, [email protected])
H. Mark Lyon – Palo Alto (+1 650-849-5307, [email protected])
Karl G. Nelson – Dallas (+1 214-698-3203, [email protected])
Deborah L. Stein (+1 213-229-7164, [email protected])
Eric D. Vandevelde – Los Angeles (+1 213-229-7186, [email protected])
Benjamin B. Wagner – Palo Alto (+1 650-849-5395, [email protected])
Michael Li-Ming Wong – San Francisco/Palo Alto (+1 415-393-8333/+1 650-849-5393, [email protected])

Europe
Ahmed Baladi – Co-Chair, PCCP Practice, Paris (+33 (0)1 56 43 13 00, [email protected])
James A. Cox – London (+44 (0)20 7071 4250, [email protected])
Patrick Doris – London (+44 (0)20 7071 4276, [email protected])
Bernard Grinspan – Paris (+33 (0)1 56 43 13 00, [email protected])
Penny Madden – London (+44 (0)20 7071 4226, [email protected])
Michael Walther – Munich (+49 89 189 33-180, [email protected])
Kai Gesing – Munich (+49 89 189 33-180, [email protected])
Alejandro Guerrero – Brussels (+32 2 554 7218, [email protected])
Vera Lukic – Paris (+33 (0)1 56 43 13 00, [email protected])
Sarah Wazen – London (+44 (0)20 7071 4203, [email protected])

Asia
Kelly Austin – Hong Kong (+852 2214 3788, [email protected])
Jai S. Pathak – Singapore (+65 6507 3683, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

The number of securities cases filed in federal court continued at a furious pace for the third year in a row. This year-end update highlights what you most need to know in securities litigation trends and developments for the last half of 2019:

Oral argument in Liu v. SEC, No. 18-1501, is scheduled for March 3, 2020, when the Supreme Court will consider the power of the SEC—and potentially, by extension, other federal agencies—to order “equitable disgorgement” in light of the Supreme Court’s prior ruling in Kokesh v. SEC, 137 S. Ct. 1635 (2017).
Anticipation for the Supreme Court’s decision in Jander—a case expected to examine the intersection of federal securities laws and ERISA—fizzled recently when the Supreme Court vacated and remanded for the Second Circuit to consider issues not resolved by its prior decision.
Developments in the Delaware Court of Chancery include continued scrutiny of relationships between directors for purposes of independence analyses, consideration of when a stockholder letter constitutes a formal demand to take corrective actions, and determining whether a buyer is excused from closing on an acquisition when the target discovers that FDA approval of its only product is at risk because of its own officer’s fraud.
Although no defendant has been found liable as a “disseminator” since the Supreme Court’s 2019 decision in Lorenzo, trial courts and the Tenth Circuit have begun to grapple with the case’s important implications.
We continue to observe Omnicare’s falsity of opinions standard developing into a formidable pleading barrier to securities fraud claims, with both the Eleventh and Fifth Circuits recently upholding dismissals at the pleadings stage.
Although the federal circuit courts of appeals did not provide any new guidance on “price impact” theories under Halliburton II during the second half of 2019, we expect the Second Circuit will soon reach a decision in Goldman Sachs II, which has been under consideration since June.
Finally, New York recently amended the statute of limitations for Martin Act claims, extending it from three years to six years.

I. Filing and Settlement Trends

Data from a newly released NERA Economic Consulting (“NERA”) study shows that 2019 was a year largely unchanged from 2018. To start, the number of new federal class action cases filed in 2019 was equal to 2018, which buttressed a trend of increased filings that began in 2017.

There has also been a continuation of the shift in the types of cases filed. The number of Rule 10b-5, Section 11 and Section 12 cases increased slightly in 2019, with 31 more filings than in 2018, while the number of merger objection cases fell.

The median settlement values of federal securities cases for 2019—excluding merger-objection cases and cases settling for more than $1 billion or $0 to the class—was roughly equivalent to those in 2018 (at $13 million, up from $12 million in 2018). However, average settlement values were down by more than 50% (at $30 million, down from $71 million in 2018). This discrepancy is due in large part to the settlement of one case in 2018 exceeding $1 billion. Excluding such an outlier, we see only a slight increase in average settlement values compared to the prior two years.

The industry sectors most frequently sued in 2019 continue to be the “Health Technology and Services” and “Electronic Technology and Technology Services” sectors, although 2019 saw the continuation of a downward trend in cases filed against healthcare companies following a spike in 2016.

A. Filing Trends

Figure 1 below reflects filing rates for 2019 (all charts courtesy of NERA). Four hundred and thirty-three cases were filed this past year, exactly matching the number of cases filed in 2018 and similar to the number of filings in 2017. However, this figure does not include the many class action suits filed in state courts or the rising number of state court derivative suits, including those filed in the Delaware Court of Chancery.

Figure 1:

B. Mix of Cases Filed in 2019

1. Filings by Industry Sector

As seen in Figure 2 below, the split of non-merger objection class actions filed in 2019 across industry sectors is fairly consistent with the distribution observed in 2018, with few indications of significant shifts or increases in particular sectors. As in 2018, the “Health Technology and Services” and the “Electronic Technology and Technology Services” sectors accounted for over 40% of filings, although there was a slight drop in “Health Technology and Services”-related filings (at 21%, down from 25% in 2018). The other two sectors reflecting the largest changes from 2018 are “Process Industries” (at 4%, up from 1% in 2018) and “Consumer and Distribution Services” (at 6%, down from 9% in 2018).

Figure 2:

2. Merger Cases

As shown in Figure 3 below, there were 170 merger objection cases filed in federal court in 2019. Although this is a 15% decrease from the number of such cases filed in 2018, the 170 filings continue the overall trend of a substantial increase in merger objection suits being filed in federal court after 2016, when the Delaware Court of Chancery put an effective end to the practice of disclosure-only settlements in In re Trulia Inc. Stockholder Litigation, 29 A.3d 884 (Del. Ch. 2016).

Figure 3:

C. Settlement Trends

As Figure 4 shows below, the average settlement value in 2019 declined by more than 50% from $71 million in 2018 to $30 million, but still remained higher than the average of $26 million in 2017. This decrease in the average settlement value can primarily be attributed to the inclusion of a settlement in 2018 that exceeded $1 billion, thereby skewing the average for that year. If our analysis is limited to cases with settlements under $1 billion, there actually is a slight increase in the average settlement value in 2019 compared to the prior years.

Figure 4:

As Figure 5 shows, the median settlement value in 2019 was $13 million, which is similar to the median in 2018 ($12 million) and almost double the median value in 2017 ($7 million).

Figure 5:

As shown in Figure 6, the Median NERA-Defined Investor Losses and Median Ratio of Actual Settlement to Investor Losses by Settlement Year remained steady in 2019 at $472 million, following a return in 2018 to a number similar to those recorded during the period 2014 through 2016.

Figure 6:

Finally, Figure 7 shows that 2018 saw increases in the percentage of settlements in the $10 to $19.9 million range, $50 to 99.9 million range, and $100+ million range. The perecentage of settlements in the $20 to $49.9 million range returned to virtually the same level that at which it was located in 2017, after experiencing a significant bump in 2018.

Figure 7:

II. What to Watch for in the Supreme Court

A. Disgorgement in SEC Enforcement Actions

On November 1, 2019, the Supreme Court granted certiorari in Charles C. Liu and Xin Wang A/K/A Lisa Wang v. SEC, No. 18-1501, to review a Ninth Circuit decision affirming summary judgment for the Securities and Exchange Commission (“SEC”) on a claim of securities fraud under Section 17(a)(2) of the Securities Act and ordering disgorgement of the entire amount that the petitioners had raised from investors.

Liu and Wang formed and controlled corporate entities presumably to build and operate a proton therapy cancer treatment center in Montebello, California. Liu financed the prospective cancer center with $27 million of international investments raised through the EB–5 Immigrant Investor Program—which allows foreigners to obtain permanent residency in the U.S. by investing at least $500,000 in a “Targeted Employment Area,” thereby creating at least 10 full-time jobs for U.S. workers.

Instead of pursuing proton therapy, Liu funneled over $20 million of investor money to himself, his wife Wang, and marketing companies associated with them. In fact, the bulk of the millions of dollars transferred occurred shortly after the SEC subpoenaed Liu as part of its initial investigation in February 2016. No permit was ever issued for the construction of the treatment center.

The SEC sought summary judgment on three securities fraud causes of action against the defendants but the district court addressed only the Section 17(a)(2) claim, given that it was a sufficient basis for the remedies sought by the SEC. See SEC v. Liu, 262 F. Supp. 3d 957, 970 (C.D. Cal. 2017). The SEC asked the court to, inter alia, order disgorgement of the total amount raised from the investors ($27 million) less the amount left over and available to be returned ($200,000). On the basis of its broad equitable power to order disgorgement of ill-gotten gains, and further discretion to indicate the amount to be disgorged, the court granted the relief sought by the SEC. See id. at 975.

On appeal to the Ninth Circuit, defendants argued that the district court’s disgorgement order was erroneous. SEC v. Liu, 754 F. App’x 505, 509 (9th Cir. 2018). Relying on Kokesh v. SEC, 137 S. Ct. 1635 (2017), defendants asserted that the district court lacked the power to order the disgorgement. Liu, 754 F. App’x at 509. In Kokesh, the Supreme Court held that disgorgement operates as a penalty, and any claim for disgorgement in an SEC enforcement action must be commenced within five years of the date the claim accrued. See Kokesh, 137 S. Ct. at 1645. Reviewing for abuse of discretion, the Ninth Circuit concluded that Kokesh expressly did not address the issue of whether a court had the equitable power to order disgorgement, thereby distinguishing it from Ninth Circuit precedent on this matter. See Liu, 754 F. App’x at 509.

In their petition for a writ of certiorari, Liu and Wang specifically questioned the equitable power to award disgorgement in the wake of Kokesh. They argued that circuit courts need guidance after Kokesh, and also challenged the use of what was characterized as “equitable disgorgement” by other agencies, including the FTC and the EPA. The Kokesh Court—in providing a historical summary of the SEC’s enforcement powers—seemed to express disapproval of the SEC’s continued use of disgorgement in enforcement proceedings. See 137 S. Ct. at 1640 (“The Act left the Commission with a full panoply of enforcement tools: It may promulgate rules, investigate violations of those rules and the securities laws generally, and seek monetary penalties and injunctive relief for those violations. In the years since the Act, however, the Commission has continued its practice of seeking disgorgement in enforcement proceedings.”).

Oral argument is scheduled for March 3, 2020. Based on the merits brief, it seems possible that the Court could issue a ruling further curtailing the SEC’s reliance on the disgorgement remedy in civil enforcement actions.

B. Intersection Between Securities Laws and ERISA

On June 3, 2019, the Supreme Court granted certiorari in Retirement Plans Committee of IBM v. Jander, No. 18-1165. Fiduciaries of the IBM retirement plan had sought review of the Second Circuit’s decision, which reversed the district court’s dismissal of retirement plan participants’ putative class complaint alleging that the Committee members breached their duties of prudence and loyalty under ERISA by continuing to invest in IBM stock while in possession of inside information about the company’s supposedly fraudulent practices.

IBM offers its employees an ERISA-qualified employee stock ownership plan (“ESOP”), which invests primarily in IBM common stock. See Jander v. Ret. Plans Comm. of IBM, 910 F.3d 620, 622 (2d Cir. 2018). Employees sued, arguing that plan fiduciaries (who were company insiders) breached their duty of prudence under ERISA by continuing to invest the plan in IBM stock despite allegedly knowing its market price was artificially inflated due to the company’s concealment of troubles in IBM’s microelectronics business. See id. at 622–23. Employees claimed that fiduciaries should either have disclosed the issues with the business’s valuation or frozen further investment in IBM stock. See id. at 623.

The district court dismissed the employees’ complaint for failure to state a claim because they failed to meet the pleading standard set forth in Fifth Third Bancorp v. Dudenhoeffer, 573 U.S. 409, 428 (2014), which requires ERISA plaintiffs to “plausibly allege an alternative action that the defendant could have taken that would have been consistent with the securities laws and that a prudent fiduciary in the same circumstances would not have viewed as more likely to harm the fund than to help it.” After employees amended their complaint to allege that disclosure of the fraud was inevitable and the harm of such disclosure would generally increase over time, as well as adding another possible alternative action fiduciaries could have taken, the district court again dismissed under Dudenhoeffer, because fiduciaries could reasonably conclude that all three alternatives would cause more harm than good. See Jander v. Ret. Plans Comm. of IBM, 272 F. Supp. 3d 444, 451–54 (S.D.N.Y. 2017).

The Second Circuit reversed, holding that “when a ‘drop in the value of the stock already held by the fund’ is inevitable, . . . it is far more plausible that a prudent fiduciary would prefer to limit the effects of the stock’s artificial inflation on the ESOP’s beneficiaries through prompt disclosure.” Jander, 910 F.3d at 630 (citation omitted) (quoting Dudenhoeffer, 573 U.S. at 430). The Second Circuit found that the employees therefore met the Dudenhoeffer standard by alleging, in part, research suggesting that the employees’ losses would have been smaller if negative information were disclosed promptly. Id. at 629–30. Plan fiduciaries petitioned for a writ of certiorari.

After certiorari was granted, the fiduciaries (and the Solicitor General, on behalf of the SEC and the Department of Labor) focused on other arguments in their merits briefing. See Ret. Plans Comm. of IBM v. Jander, 2020 WL 201024, at *1 (U.S. Jan. 14, 2020) (per curiam). The fiduciaries argued for a bright-line rule that ERISA could never impose a duty on them to act on inside information. Id. The Government argued that requiring fiduciaries to disclose inside information under ERISA that is not otherwise required to be disclosed by the securities laws would conflict with the complex disclosure requirements imposed by those laws.

Rather than resolve the questions presented on the pleading standard in breach of fiduciary duty cases involving employee-benefit plans, on January 14, 2020 the Supreme Court vacated and remanded back to the Second Circuit for consideration of the issues raised in the merits briefing that were not resolved by the previous decision. Id. at *2. In remanding, the Court referenced its statement in Dudenhoeffer that the SEC’s views “might ‘well be relevant’ to discerning the content of ERISA’s duty of prudence in this context.” Id. (quoting Dudenhoeffer, 573 U.S. at 429).

Justice Kagan authored a concurring opinion, joined by Justice Ginsburg, noting that the Second Circuit could refuse to hear these new arguments if they were not properly preserved. Id. (Kagan, J., concurring). And if the Second Circuit did choose to address them, Justice Kagan opined that they would be hard to square with Dudenhoeffer, as that case “makes clear that an ESOP fiduciary at times has . . . a duty” to act on insider information given that it “sets out exactly what a plaintiff must allege to state a claim that the fiduciary breached his duty of prudence by ‘failing to act on inside information.’” Id. (quoting Dudenhoeffer, 573 U.S. at 423). Justice Kagan disagreed with the Government’s argument that ERISA only imposes such a duty when already imposed by the securities laws, explaining that Dudenhoeffer only holds that there is no duty to disclose when it would “violat[e],” or “conflict[]” with the “requirements” or “objectives” of those laws. Id. (quoting Dudenhoeffer, 573 U.S. at 428–29). Justice Kagan therefore left open the possibility that disclosure might be required under ERISA “even if the securities laws do not require it,” positing that in such a “conflict-free zone” the question would be whether a “prudent fiduciary would think the action more likely to help than to harm the fund.” Id. (citing Dudenhoeffer, 573 U.S. at 428).

Justice Gorsuch also authored a concurring opinion, disagreeing with Justice Kagan’s “broad[]” reading of Dudenhoeffer, and noting that the “pure question of law” raised in the case should be “addressed immediately.” Id. at *3 (Gorsuch, J., concurring). Under Justice Gorsuch’s view, Dudenhoeffer does not impose liability on plan fiduciaries for “alternative actions they could have taken only in a nonfiduciary capacity.” Id.

As Jander involves important questions regarding the fiduciary duties of pension plan managers who invest in company stock, including the intersection between the securities laws and ERISA, readers can expect that this will not be the Supreme Court’s last word on the issue.

III. Delaware Developments

A. The Delaware Court of Chancery Continues to Scrutinize Relationships Between Directors

Over the last several years, Delaware courts have reviewed independence among directors with seemingly increased scrutiny. See, e.g., Marchand v. Barnhill, 212 A.3d 805 (Del. 2019) (director’s 28-year relationship with CEO’s family rebutted presumption of independence); Sandys v. Pincus, 152 A.3d 124 (Del. 2016) (director’s 50-year friendship with controller rebutted presumption of independence); Del. Cty. Emps. Ret. Fund v. Sanchez, 124 A.3d 1017 (Del. 2015) (director and controller’s co-ownership of airplane rebutted presumption of independence). The Court of Chancery continued that apparent trend in In re BGC Partners, Inc., where it closely scrutinized the relationships between the members of BGC’s board of directors and the company’s controller. 2019 WL 4745121, at *1 (Del. Ch. Sept. 30, 2019).

In BGC, stockholders of BGC purported to bring a derivative action against its controlling stockholder—who also served as its Chairman and CEO—and other directors based on the theory that the controller caused BGC to acquire and overpay for another company in which the controller owned a controlling stake. Id. at *1–2. As the controller’s interest in the transaction was conceded, the court’s analysis of whether the plaintiffs adequately pleaded demand futility and rebutted the business judgment rule turned on whether a majority of BGC’s directors were interested in the proposed transaction or lacked independence with respect to the controller. Id. at *9.

Based on a deep dive into three particular directors’ professional and personal connections to the controller, id. at *10–14, the court held that it could infer that a majority of directors lacked independence, id. at *9. For the first director, the court noted that he had a 20-year professional and personal relationship with the controller, including attending galas with each other’s families and the controller’s setting up a private tour of a museum for the director’s family. Id. at *11–12. For the second and third directors, the court focused on their service on the boards of other companies affiliated with the controller and how their income from this service was likely to be material relative to their other sources of income. Id. at *12–14. The Court also referenced both of these directors’ ties to a college to which the controller had made substantial donations. Id. Although one director was no longer affiliated with that college, the court explained that past benefits could be enough to create a sense of obligation to the controller. Id. at *12.

B. Court of Chancery Interprets Demand Letter

Whether a stockholder’s letter to the board is a “demand” affects the standard of review applicable to any litigation arising from that letter. If the letter is indeed a demand, then, under Delaware law, the stockholder has “tacitly concede[d]” that the board was able to exercise its business judgment in considering it. Spiegel v. Buntrock, 571 A.2d 767, 777 (Del. 1990). In Solak v. Welch, 2019 WL 5588877 (Del. Ch. Oct. 30, 2019), the Delaware Court of Chancery held that a stockholder’s letter was a “demand” even though it did not expressly demand litigation.

The stockholder plaintiff in Solak sent a letter to the company’s board of directors “to suggest that the [board] take corrective action to address excessive director compensation as well as compensation practices and policies pertaining to directors.” Solak, 2019 WL 5588877, at *2. The letter asserted that the company’s compensation policy “lacks any meaningful limitations” and “warn[ed]” that “[t]he company is more susceptible than ever to shareholder challenges unless it revises or amends its director compensation practices and policies.” Id. The letter “suggest[ed]” that the board “take immediate remedial measures” and stated that the plaintiff “would consider ‘all available stockholder remedies’” if the board failed to respond within 30 days. Id. But the letter also included a footnote saying that “nothing contained herein shall be construed as a pre-suit litigation demand under Delaware Chancery Rule 23.1,” and that “[w]e do not seek or expect the board to initiate any legal action against its members.” Id.

The board sent a response letter explaining that it viewed the stockholder letter as a demand, and declined to take any of the remedial actions suggested in the stockholder letter. Id. at *3. So the stockholder sued, purporting to assert derivative claims. Id. At issue was whether the letter counted as a “demand” on the board. Id. at *4. The court explained that a pre-suit communication need not expressly demand litigation to be deemed a demand. Id. at *5–6. Rather, the letter need only “clearly articulat[e] the remedial action to be taken by the board” or “clearly demand[] corporate action.” Id. at *5. The letter’s “strong overtures of litigation” and suggested remedial measures met this test, notwithstanding its footnote purportedly disclaiming that it was a demand. Id. at *6–7. And because the letter was a demand, the strict demand-refused standard applied, which the plaintiff could not overcome. Id. at *8–9.

**C. Despite Akorn, an MAE Is Still a Rare Event Requiring a Buyer to Carry a Heavy Burden**

As we discussed in our 2018 Year-End Securities Litigation Update, in 2018, the Delaware Supreme Court affirmed the Court of Chancery’s conclusion that a buyer had proven it properly terminated a merger agreement because the target had suffered a “material adverse effect” (or “MAE”)—a first for both courts. See Akorn, Inc. v. Fresenius Kabi AG, 2018 WL 4719347 (Del. Ch. Oct. 1, 2018), aff’d, 198 A.3d 724 (Del. 2018). As the Court of Chancery explained, the test for an MAE is “whether there has been an adverse change in the target’s business that is consequential to the company’s long-term earnings power over a reasonable period, which one would expect to be measured in years rather than months.” Id. at *53.

In Channel Medsystems, the first case since Akorn to consider whether an MAE had occurred, the Court of Chancery confirmed that triggering an MAE clause remains a high bar. Channel Medsystems, Inc. v. Boston Scientific Corp., 2019 WL 6896462 (Del. Ch. Dec. 18, 2019). In that case, Boston Scientific sought to be relieved from its agreement to acquire Channel after Channel learned and disclosed that fraud committed by its Vice President of Quality put at risk FDA approval of its only device even though “the FDA [had] accepted Channel’s remediation plan” and “made the FDA’s approval a distinct possibility.” Id. at *1. Indeed, one month before trial in Channel Medsystems, and “consistent with the timeframe for receiving FDA approval the parties expected when they entered into the [merger agreement],” the FDA approved the device. Id.

The Court of Chancery concluded that “Boston Scientific failed to prove based on both qualitative and quantitative factors that it was entitled to terminate [the parties’ agreement].” Id. at *36.

First, the court considered whether Boston Scientific held, at the time it purported to terminate the deal, “a reasonable expectation” that Channel “would reasonably be expected” to suffer a qualitatively significant adverse effect as of the closing date. Id. at *25, *29. But the court found virtually no contemporaneous evidence suggesting that Boston Scientific held such an expectation. Id. at *33. To the contrary, Boston Scientific had failed to take any reasonable steps to make an informed decision regarding the likely impact of the fraud on Channel and instead had relied solely on a report provided by Channel, which actually concluded the fraud had no impact on the device. Id. at *29–30.

Second, the court rejected Boston Scientific’s attempt to demonstrate the quantitative impact of the fraud on Channel’s value to Boston Scientific as of the date that the merger agreement was signed. Id. at *34. The court did so in large part because Boston Scientific based its expert’s analysis on assumptions that were not objectively reasonable. Id. In particular, Boston Scientific’s expert assumed that Channel’s only product would have to be held off the market for two to four years for remediation and retesting. Id. The court found that this assumption was not objectively reasonable, however, because “Boston Scientific’s own track record and the testimony of its own witnesses belie[d] the contention that it was necessary to remediate an retest [the device] before placing it on the market given the FDA’s approval of the device.” Id. at *28–29, 35.

IV. Lower Courts Grappling with Implications of Lorenzo

As we discussed in our 2019 Mid-Year Securities Litigation Update, on March 27, 2019, the Supreme Court held in Lorenzo v. SEC, 139 S. Ct. 1094 (2019), that those who disseminate false or misleading information to the investing public with the intent to defraud can be found liable under Section 17(a)(1) of the Securities Act and under Exchange Act Rules 10b-5(a) and 10b-5(c), even if the disseminator did not “make” the statements and was thus not subject to enforcement under Rule 10b-5(b).

Importantly, in Lorenzo, the Court stated that “[t]hose who disseminate false statements with intent to defraud are primarily liable under Rules 10b-5(a) and (c),” as well as Section 10(b) of the Exchange Act and Section 17(a)(1) of the Securities Act, “even if they are secondarily liable under Rule 10b-5(b).” Lorenzo, 139 S. Ct. at 1104. This holding raises the possibility that secondary actors could face liability under Exchange Act Rules 10b-5(a) and 10b-5(c) simply for disseminating the alleged misstatement of another if a plaintiff can show that they knew the statements contained false or misleading information. Although this issue has yet to come up in other cases, over the last year, three lower federal courts have grappled with how to apply Lorenzo in other ways.

First, in April 2019, the Southern District of New York relied on Lorenzo to find that the SEC had adequately pleaded scheme liability under Rule 10b-5(a) and (c) even though it had alleged no deceptive act other than misstatements or omissions. SEC v. SeeThruEquity, LLC, 2019 WL 1998027 (S.D.N.Y. Apr. 26, 2019). The defendants, a stock research company and its co-founders, were accused of failing to disclose that they were paid by a company that they were recommending in their research reports. Id. at *1–3. They argued the SEC could not plead “scheme liability” under Rule 10b-5 because they had been accused of no deceptive acts beyond the misstatements themselves. Id. at *5. The court rejected this argument, stating that “[t]he complaint alleges that the defendants’ entire business model, beyond any misstatements or omissions, is deceptive.” Id.

Then, in August 2019, the Tenth Circuit expanded Lorenzo further, holding that scheme liability could be found based on a failure to correct a misstatement. See Malouf v. SEC, 933 F.3d 1248 (10th Cir. 2019). In Malouf, the defendant had occupied two key positions at separate firms—one was a branch of the broker-dealer firm Raymond James Financial Services (“Raymond James”) and the other, UASNM, Inc. (“UASNM”), provided clients with investment advice. Id. at 1253–54. After Raymond James became concerned about the defendant’s dual role at the two firms, the defendant sold his Raymond James branch, which was to be paid for in installments based on the branch’s “collection of securities-related fees.” Id. at 1254. To collect on the sale, the defendant routed bond trades for his UASNM clients through the Raymond James branch so that the branch’s buyer could pay the defendant back with money accrued through commissions. Id. The defendant did not disclose this arrangement to anyone at UASNM, which publicly touted that it provided its clients with “impartial advice untainted by any conflicts of interest.” Id. Meanwhile, the defendant also helped decide what UASNM would include in its public disclosures, but “took no steps to remedy UASNM’s misstatements or to disclose his own conflict of interest.” Id. at 1254–55. Ultimately, after an outside consultant caught wind of the conflict, it was disclosed. Id. at 1255. During an enforcement action, the administrative law judge found that the defendant had violated, among other things, Section 17(a)(1) of the Securities Act and Exchange Act Rules 10b-5(a) and 10b-5(c). Id.

On appeal, the Tenth Circuit affirmed. Id. at 1253. In connection with Section 17(a)(1) of the Securities Act and Exchange Act Rules 10b-5(a) and 10b-5(c), the court stated “we conclude that [the defendant’s] failure to correct UASNM’s misstatements could trigger liability” because, under Lorenzo, “a person could incur liability under these provisions when the conduct involves another person’s false or misleading statement.” Id. at 1259–60 (citing 139 S. Ct. at 1101–03). In other words, the panel accepted that the defendant was liable because, although he did not disseminate UASNM’s alleged misstatements, he failed to correct the relevant disclosures that he knew were false.

Finally, in December 2019, in EnSource Investments LLC v. Willis, 2019 WL 6700403 (S.D. Cal. Dec. 6, 2019), a court found that Lorenzo did not apply to entities involved in an allegedly fraudulent scheme because those entities had not “disseminated any false statements.” Id. at *13. In EnSource, two entities were “under the umbrella” of another company and its founder, both of whom were defendants in the case. Id. at *1. The founder and parent company were found to have made misstatements “on behalf” of the entities. Id. at *13. Rather than holding that these entities had a duty under Lorenzo to correct the misstatements made on their behalf, the EnSource court simply found that because the entities did not disseminate the misstatements, Lorenzo did not apply. Id. at *13.

It remains to be seen whether cases such as SeeThruEquity or Malouf will be confined to their facts or whether courts will adopt or expand on these holdings to increase the reach of scheme liability. We will, of course, provide an update on the direction that courts take Lorenzo and scheme liability in our 2020 Mid-Year Securities Litigation Update.

V. Falsity of Opinions – Omnicare Update

As we discussed in our prior securities litigation updates, lower courts continue to explore application of the standard set forth in Omnicare, Inc. v. Laborers District Council Construction Industry Pension Fund, 575 U.S. 175 (2015), for determining falsity of an opinion. In its Omnicare decision, the Supreme Court addressed the scope of liability for false opinion statements under Section 11 of the Securities Act and held that “a sincere statement of pure opinion is not an ‘untrue statement of material fact,’ regardless whether an investor can ultimately prove the belief wrong.” Id. at 186. According to that standard, an opinion statement can give rise to liability only when the speaker does not “actually hold[] the stated belief,” or when the opinion statement contains “embedded statements of fact” that are untrue. Id. at 184–85. In the “omission” section of the opinion, the Court held that a factual omission “about the issuer’s inquiry into or knowledge concerning a statement of opinion” gives rise to liability when the omitted facts “conflict with what a reasonable investor would take from the statement itself.” Id. at 1329.

Omnicare’s falsity of opinions standard continues to serve as a significant pleading barrier to securities fraud claims. In Carvelli v. Ocwen Financial Corp., the Eleventh Circuit Court of Appeals held that the plaintiff failed to show the defendant’s “statements of opinion are mutually exclusive of—or even inconsistent with—[the company]’s alleged knowledge,” and therefore the complaint failed to meet the pleading standard set forth in Omnicare. 934 F.3d 1307, 1323 (11th Cir. 2019). The court noted that merely an inference that the company “could or should have known” that its belief about the company’s economic vitality conflicted with the company’s “persistent” technology problems is insufficient to show the company did not believe its statements of opinion. Id. (emphasis original).

In its first decision applying the standard for opinion liability post-Omnicare, the Fifth Circuit Court of Appeals affirmed the dismissal of a case concerning an oil company’s stated belief that it was in “substantial compliance” with regulatory obligations. Police & Fire Ret. Sys. of City of Detroit v. Plains All Am. Pipeline, L.P., 777 F. App’x 726 (5th Cir. 2019) (“Plains II”). As discussed in our 2018 Mid-Year Securities Litigation Update, the Southern District of Texas dismissed allegations that statements concerning compliance were misleading on the basis that a regulatory agency had identified issues that concerned only a different and small part of the company’s varied operations. In re Plains All Am. Pipeline, L.P. Sec. Litig., 307 F. Supp. 3d 583, 621–22 (S.D. Tex. 2018). The Fifth Circuit concurred that the company’s “belief statements” regarding its compliance “were broadly applicable and therefore were not rendered false or misleading” by issues that affected “a small percentage” of the company’s pipelines. Plains II, 777 F. App’x at 731.

In the latter half of 2019, several courts reached differing conclusions on whether companies could be held liable for opinions about the results of scientific research. In Lehman v. Ohr Pharmaceuticals, plaintiffs alleged that a company’s optimistic announcements about second-phase drug trials were misleading where the company omitted that the results were only meaningful because the control group fared significantly worse than in historical trials. 2019 WL 4572765 (S.D.N.Y. Sept. 20, 2019). The Southern District of New York disagreed, relying on the Second Circuit’s opinion in Tongue v. Sanofi, in which the court found that a pharmaceutical company’s statements were not misleading even though they did not “include a fact that would have potentially undermined Defendants’ optimistic projections.” Id. at *3 (citing Tongue v. Sanofi, 816 F.3d 199, 212 (2d Cir. 2016)). Judge Preska also cautioned against courts issuing decisions that would compel caution rather than optimism about the results of such an experiment: “[T]he law does not abide attempts at using the judiciary to stifle the risk-taking that undergirds scientific advancement and human progress. The answer to bad science is more science, not this Court’s acting as the Southern District for the Inquisition.” Id. at *5.

By contrast, in Micholle v. Ophthotech Corp., the court considered whether an opinion that a change in testing methodology had no “meaningful” impact on who was eligible to participate in a certain drug trial was actionable in light of plaintiff’s allegations that there was at least a 17% difference. 2019 WL 4464802, at *12 (S.D.N.Y. Sept. 17, 2019). The court denied dismissal because “[m]ateriality is a fact-specific inquiry” and an “investor may well have considered the degree of similarity between the parameters of a new clinical trial and those of a recently completed—and purportedly very successful—clinical trial important.” Id. at *13.

There were a handful of reported decisions that focused on whether a complaint sufficiently pled the omission of contrary facts that rendered positive opinions regarding the company’s business misleading. For example, in Hawaii Structural Ironworkers Pension Trust Fund v. AMC Entertainment Holdings, Inc., plaintiffs plausibly alleged that opinions about the “smooth” process of integrating a recent acquisition implied “that there were no significant or systemic obstacles to [the] integration.” 2019 WL 4601644, at *12 (S.D.N.Y. Sept. 23, 2019). Similarly, in Vignola v. FAT Brands, Inc., a Central District of California court denied the defendants’ motion to dismiss statements concerning the experience and track record of the company’s senior leadership team. 2019 WL 6888051, at *10 (C.D. Cal. Dec. 17, 2019). The court considered that while investors do understand that opinions generally are formed by weighing competing facts, here, the company allegedly omitted the key fact “that the same leadership team had previously steered the subsidiaries of its same flagship brand into bankruptcy.” Id. at *10 (emphasis original).

VI. Halliburton II Market Efficiency and “Price Impact” Cases

We are continuing to monitor significant decisions interpreting Halliburton Co. v. Erica P. John Fund, Inc., 573 U.S. 258 (2014) (“Halliburton II”). The federal circuit courts of appeals did not provide any new guidance in the second half of 2019, but certain questions have been recurring in trial courts recently. Recall that in Halliburton II, the Supreme Court preserved the “fraud-on-the-market” presumption, permitting plaintiffs to maintain the common proof of reliance that is required for class certification in a Rule 10b-5 case, but also permitting defendants to rebut the presumption at the class certification stage with evidence that the alleged misrepresentation did not impact the issuer’s stock price. The key questions we have been following in the wake of Halliburton II are the following: (1) How should courts reconcile the Supreme Court’s explicit ruling in Halliburton II that direct and indirect evidence of price impact must be considered at the class certification stage, Halliburton II, 573 U.S. at 283, with the Supreme Court’s previous decisions in Erica P. John Fund, Inc. v. Halliburton Co., 563 U.S. 804 (2011) (“Halliburton I”), and Amgen Inc. v. Connecticut Retirement Plans & Trust Funds, 568 U.S. 455 (2013), holding that plaintiffs need not prove loss causation or materiality until the merits stage?; (2) What standard of proof must defendants meet to rebut the presumption with evidence of no price impact?; and (3) What evidence is required to successfully rebut the presumption?

As previously discussed in our 2018 Year-End Securities Litigation Update, the Second Circuit addressed the first two questions in Waggoner v. Barclays PLC, 875 F.3d 79 (2d Cir. 2017) (“Barclays”) and Arkansas Teachers Retirement System v. Goldman Sachs, 879 F.3d 474 (2d Cir. 2018) (“Goldman Sachs”). Those decisions remain the most substantive interpretations of Halliburton II. Barclays held that after a plaintiff establishes the presumption of reliance applies, the defendant bears the burden of persuasion to rebut the presumption by a preponderance of the evidence. Barclays, 875 F.3d at 100–03. Though this appeared to put the Second Circuit at odds with the Eighth Circuit, which cited Rule 301 of the Federal Rules of Evidence when reversing a trial court’s certification order on price impact grounds, IBEW Local 98 Pension Fund v. Best Buy Co., 818 F.3d 775, 782 (8th Cir. 2016), the inconsistency was not enough to persuade the Supreme Court to take up the issue, Barclays PLC v. Waggoner, 138 S. Ct. 1702 (Mem.) (2018) (denying writ of certiorari).

In Goldman Sachs, the Second Circuit vacated the trial court’s ruling certifying a class and remanded the action, directing that price impact evidence must be analyzed prior to certification, even if price impact “touches” on the issue of materiality. Goldman Sachs, 879 F.3d at 486. Recent district court decisions in the latter half of 2019 have embraced this approach when reconciling Halliburton II with Halliburton I and Amgen. See, e.g., In re Chicago Bridge & Iron Co. N.V. Sec. Litig., 2019 WL 5287980, at *21–23 (S.D.N.Y. Oct. 18, 2019) (concluding price impact analysis appropriate prior to class certification even if it may “touch on materiality”); Di Donato v. Insys Therapeutics, Inc., 2019 WL 4573443, at *6 (D. Ariz. Sept. 20, 2019) (explaining that a plaintiff need not prove materiality at the class certification stage). The Southern District of New York again certified the Goldman Sachs class, In re Goldman Sachs Grp. Sec. Litig., 2018 WL 3854757, at *1–2 (S.D.N.Y. Aug. 14, 2018), holding that while the price had not moved in response to previous statements on the same subject as the alleged corrective disclosures, those disclosures were sufficiently different to credit plaintiff’s expert’s “link between the news of [defendant]’s conflicts and the subsequent stock price declines,” and that defendants’ expert testimony was insufficient to “sever” that link. Id. at *4–6. The Second Circuit agreed to review the decision recertifying the class, see Order, Ark. Teachers Ret. Sys. v. Goldman Sachs, Case No. 18-3667 (2d Cir. Jan. 31, 2019) (“Goldman Sachs II”), and the case was fully briefed, argued and taken under consideration in June. A decision could be reached in the case any day now.

In 2019, the Third Circuit also weighed in, providing some guidance on the type of evidence defendants must present to rebut the presumption of reliance at the class certification stage. That court affirmed the district court’s grant of plaintiff’s motion for certification, finding that the district court did not abuse its discretion in considering conflicting expert testimony. Vizirgianakis v. Aeterna Zentaris, Inc., 775 F. App’x 51, 53–54 (3d Cir. 2019). Most significantly, the Third Circuit rejected defendants’ argument that plaintiff’s expert’s event study, which was offered for the purpose of proving market efficiency (i.e., that the stock price moved in reaction to news about the company), was actually evidence that the statements at issue had no price impact. Id. at 53. Specifically, Defendants argued that because plaintiff’s expert had not proven a stock price movement in response to one of the alleged corrective disclosures at the statistically significant 95% confidence level, the relevant statement had no price impact. Id. at 53. The Court observed that plaintiff’s expert’s report was not written for the purpose of proving or disproving price impact and plaintiff’s inconclusive evidence regarding a stock price movement is not evidence of a lack of price impact. Id. at *53. Similar attempts to use plaintiffs’ market efficiency studies as evidence of a lack of price impact have been rejected by a number of district courts as well. See, e.g., In re Signet Jewelers Ltd. Sec. Litig., 2019 WL 3001084, at *13–15 (S.D.N.Y. July 10, 2019) (“Defendants’ failure to . . . supplement [their expert’s] report with an event study showing the absence of price impact is, on its own, a basis for rejecting Defendants’ arguments.”); Di Donato, 2019 WL 4573443, at *13 (“The lack of statistically significant proof that a statement affected the stock price is not a statistically significant proof of the opposite[.]”) (emphasis original); accord Chicago Bridge, 2019 WL 5287980, at *12–14 (noting that even statistically insignificant findings of causes of price impacts should be considered, albeit possibly granted less weight than statistically significant findings).

These cases suggest defendants should consider performing their own event studies if defendants intend to argue a lack of price impact rather than simply criticizing event studies offered by plaintiffs. Defendants should also account for the precise facts and circumstances of each case before settling on a strategy for challenging price impact in whole or in part.

We will continue to monitor developments in Goldman Sachs II and related cases.

VII. Statute of Limitations for Martin Act Claims Extended to Six Years

On August 26, 2019, New York Governor Andrew Cuomo signed into law a bill extending the statute of limitations for all claims brought pursuant to the Martin Act, New York’s blue sky law, to six years. This reverses a 2018 decision from New York’s highest court, which held that many Martin Act claims must be brought within three years. See generally Schneiderman v. Credit Suisse Sec. (USA) LLC, 31 N.Y.3d 622 (2018). According to the bill’s sponsor memo, a “six-year timeline was essential to some of the most meaningful cases that have reined in Wall Street excesses, halted fraudulent practices, and returned millions of dollars to defrauded consumers and investors,” and the law’s drafters expect this new six-year period to give New York’s Attorney General time to make “extensive investigations” into “novel areas of business practices.” See NY State Senate Bill S6536, The New York State Senate, https://www.nysenate.gov/legislation/bills/2019/s6536.

As readers will know, the Martin Act permits New York “to investigate and enjoin fraudulent practices in the marketing of stocks, bonds and other securities within or from New York State.” Credit Suisse, 31 N.Y.3d at 629. The Martin Act defines “fraudulent practices” “expansive[ly],” and “prohibitions against fraud, misrepresentation and material omission are found throughout the statutory scheme.” Id.; N.Y. Gen. Bus. L. §§ 352–353. Moreover, unlike common law fraud, Martin Act liability does not require any showing of “scienter or justifiable reliance on the part of investors.” Credit Suisse, 31 N.Y.3d at 632. A violation of the Martin Act can lead to both civil and criminal liability. N.Y. Gen. Bus. L. §§ 353, 358.

In 2018, in Credit Suisse, the New York Court of Appeals held that different theories of Martin Act liability would be subject to different limitations periods. When a case was premised on a legal theory akin to “fraud recognized in the common law,” the applicable statute of limitations would be six years, but for liability premised on the more “expansive” notions of a “fraudulent practice” solely created by statute, the applicable limitations period would be three years. See Credit Suisse, 31 N.Y.3d at 633–34 (dismissing stale claims).

Credit Suisse now has been expressly superseded. It is too early to predict the practical impact of this development. However, it is reasonable to assume that New York enforcement officials will quickly take advantage of the longer limitations period. In a joint statement with New York Attorney General Letitia James, Governor Cuomo trumpeted the new law as “enhancing one of the state’s most powerful tools to prosecute financial fraud so we can hold more bad actors accountable, protect investors and achieve a fairer New York for all.” New Law Strengthens AG James’ Authority To Take On Corporate Misconduct, New York State Attorney General (Aug. 26, 2019), https://ag.ny.gov/press-release/2019/new-law-strengthens-ag-james-authority-take-corporate-misconduct. Attorney General James stated that “[a]s the federal government continues to abdicate its role of protecting investors and consumers, this law is particularly important. New York remains committed to finding and prosecuting the bad actors that rob victims and destabilize markets.” Governor Cuomo Signs Legislation Increasing New York’s Capacity to Prosecute Financial Fraud, Official Website of the State of New York (Aug. 26, 2019), https://www.governor.ny.gov/news/governor-cuomo-signs-legislation-increasing-new-yorks-capacity-prosecute-financial-fraud.

The following Gibson Dunn lawyers assisted in the preparation of this client update: Mark Perry, Monica Loseman, Brian Lutz, Jefferson Bell, Shireen Barday, Nancy Hart, Lissa Percopo, Mark Mixon, Zoey Goldnick, Jason Hilborn, Hannah Kirshner, Emily Riff, Andrew Bernstein, Tim Deal, Luke Dougherty, Marc Aaron Takagaki, Patrick Taqui, Michael Klurfeld, Alisha Siqueira, Collin Vierra, and Chase Weidner.

Gibson Dunn lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work, or any of the following members of the Securities Litigation practice group steering committee:

Brian M. Lutz – Co-Chair, San Francisco/New York (+1 415-393-8379/+1 212-351-3881, [email protected])
Robert F. Serio – Co-Chair, New York (+1 212-351-3917, [email protected])
Meryl L. Young – Co-Chair, Orange County (+1 949-451-4229, [email protected])
Jefferson Bell – New York (+1 212-351-2395, [email protected])
Jennifer L. Conn – New York (+1 212-351-4086, [email protected])
Thad A. Davis – San Francisco (+1 415-393-8251, [email protected])
Ethan Dettmer – San Francisco (+1 415-393-8292, [email protected])
Barry R. Goldsmith – New York (+1 212-351-2440, [email protected])
Mark A. Kirsch – New York (+1 212-351-2662, [email protected])
Gabrielle Levin – New York (+1 212-351-3901, [email protected])
Monica K. Loseman – Denver (+1 303-298-5784, [email protected])
Jason J. Mendro – Washington, D.C. (+1 202-887-3726, [email protected])
Alex Mircheff – Los Angeles (+1 213-229-7307, [email protected])
Robert C. Walters – Dallas (+1 214-698-3114, [email protected])
Aric H. Wu – New York (+1 212-351-3820, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

On January 30, 2020, the Board of Governors of the Federal Reserve System issued a final rule that would update and revise, to some degree, its framework for finding “control” under the Bank Holding Company Act of 1956, as amended (BHC Act).

The new control rule (Control Rule) expands the relationships that an investor can have with a target institution and still be deemed to be non-controlling under the BHC Act. This is relevant both for investments in banking organizations, such as by private equity investors, and for investments by banking organizations, such as in fintech companies.

The Control Rule has the most benefits for investors below 10 percent voting share ownership, as will be described more fully below. In addition to benefiting from broader consent rights and greater business relationships than previously, such investors also have greater power as shareholders to make use of proxy solicitation to challenge management. At the same time, certain aspects of the Control Rule, such as its approach to calculating total equity of a target company, were not expanded from the original proposal and may hinder new investments.

This Client Alert describes the most significant aspects of the Control Rule, which will be effective on April 1, 2020.

I. Statutory Control Under the Bank Holding Company Act

The BHC Act defines “control” in the following manner:

a company (First Company) directly or indirectly owns, controls, or has power to vote 25 per centum or more of any class of voting securities of a company (Second Company);
the First Company controls in any manner the election of a majority of the directors or trustees, or persons performing similar functions, of the Second Company; or
the Federal Reserve determines, after notice and opportunity for hearing, that the First Company directly or indirectly exercises a controlling influence over the management or policies of the Second Company

The Federal Reserve has not held a control proceeding in decades. Instead, it has developed what may be called a “common law of control,” through individual interpretations and policy statements, in which it has set forth the factors that govern whether a “controlling influence” would exist in particular investments below 25% voting share ownership. It is this “common law of control” that the Control Rule codifies, and, in some cases expands.

II. New Presumption of Non-Control – 9.99% Rather Than 4.99% Voting Shares

Historically, the Federal Reserve presumed that a company did not have control over another company if it owned less than 5 percent of all classes of that company’s voting shares. The Control Rule introduces a new presumption. It states that the Federal Reserve will presume that a First Company does not control a Second Company if it owns 9.99% or less of all classes of the Second Company’s voting shares and no other presumptions of control exist. The presumptions of control differ depending on whether voting share ownership is between 0% and 4.99%, or between 5% and 9.99%.

A. Significantly Expanded Rights for Investors at 4.99% Voting Shares or Less

The Control Rule addresses an inconsistency in the Federal Reserve’s old control interpretations. Under them, a lender with no equity could impose covenants that could substantially restrict the ability of a bank or company to conduct its business without being deemed to control that company. Notwithstanding the presumption of non-control at 4.99% voting shares, an equity investor with de minimis equity ownership could not impose the same restrictions by contract. The Federal Reserve has eliminated this inconsistency, to the benefit of 4.99% or less investors, as described immediately below. Importantly, the Federal Reserve also clarified that the expanded rights for 4.99% or less investors apply both in the case of financial investments as well as nonfinancial investments made under Section 4(c)(6) of the BHC Act.

B. Presumptions of Control

The Control Rule sets forth differing presumptions of control, depending on the percentage of a class of voting shares in the Second Company that a First Company holds. A summary chart of these presumptions is attached as Appendix 1, and they are as follows (although the Federal Reserve expresses them as presumptions, they act as limitations, so we will describe them in the manner they function):

Control Limitations at 4.99% Voting Share Ownership or Less

The First Company may not appoint 50 percent of more of the Second Company’s directors
The First Company may own up to, but not including, one-third of the total equity of the Second Company[1]
There are no limitations on director service as Board Chair or on Board committees
There are no limitations on solicitation of proxies
All senior management interlocks are permitted
The First Company may have contractual rights that limit the Second Company’s business, as long as they do not constitute a “management agreement”
The First Company may have any amount of business relationships with the Second Company, and they are not required to be on market terms

Control Limitations at 5% – 9.99% Voting Share Ownership

The First Company may not appoint a quarter or more of the Second Company’s directors
The First Company may own up to, but not including, one-third of the total equity of the Second Company
There are no limitations on director service as board Chair or on board committees
There are no limitations on solicitation of proxies
Only one senior management interlock is permitted, not the CEO
The First Company may not have contractual rights that significantly restrict discretion
Business relationships between the two companies must generate less than 10% of (i) the total GAAP-consolidated revenues and (ii) total GAAP-consolidated expenses of the Second Company; they are not required to be on market terms[2]

Control Limitations at 10% – 14.99% Voting Share Ownership

The First Company may not appoint a quarter or more of the Second Company’s directors
The First Company may own up to, but not including, one-third of the total equity of the Second Company
There are no limitations on director service as board Chair; the First Company may have a quarter or less representation on those board committees having power to bind the organization; otherwise no limitations on committee service
Proxy solicitations are permitted, except those to replace a quarter of the board of directors or more
Only one senior management interlock is permitted, not the CEO
The First Company may not have contractual rights that significantly restrict discretion
Business relationships between the two companies must generate less than 5% of (i) the total GAAP-consolidated revenues and (ii) total GAAP-consolidated expenses of the Second Company; they are required to be on market terms

Control Limitations at 15 – 24.99% Voting Shares Ownership

The First Company may not appoint a quarter or more of the Second Company’s directors
The First Company may own up to, but not including, one-third of the total equity of the Second Company
A First Company director may not be the board Chair of the Second Company; the First Company may have a quarter or less representation on those board committees having power to bind the organization; otherwise no limitations on committee service
Proxy solicitations are permitted, except those to replace a quarter of the board of directors or more
No senior management interlocks are permitted
The First Company may not have contractual rights that significantly restrict discretion
Business relationships between the two companies must generate less than 2% of (i) the total GAAP-consolidated revenues and (ii) total GAAP-consolidated expenses of the Second Company; they are required to be on market terms

C. “Management Agreements” Versus “Rights That Significantly Restrict Discretion”

A 4.99% or less voting share investor is able to exercise the greatest amount of influence over a target bank or company. The only limitation on such investors is that they may not have “management agreements” with the Second Company. Such agreements give significant influence over the “general management or overall operations of the Second Company;” the Control Rule gives as an example the rights of a general partner over a partnership or managing member over a limited liability company.

Investors with between 5% and 24.99%, by contrast, are more restricted: they may not have contractual rights that significantly restrict discretion. The Control Rule provides examples of rights that do, and do not, act in this matter.

Rights that significantly restrict discretion include:

Prohibitions on entering into new lines of business, making substantial changes to or discontinuing new lines of business
Restricting the Second Company’s ability to merge or consolidate, or its ability to acquire, sell, lease, transfer, spin-off, recapitalize, liquidate, dissolve or dispose of subsidiaries or assets
Restricting the Second Company’s payment of dividends
Restricting the Second Company’s ability to authorize or issue additional junior debt or equity securities
Restricting the Second Company’s ability to engage in a public offering or to list or de-list securities on an exchange, other than a right that allows the securities of the First Company to have the same status as other securities of the same class
Restricting the removal or selection of any independent accountant, auditor, investment adviser or investment banker by the Second Company
Restricting the Second Company’s ability to significantly alter accounting methods and policies

Rights that do not significantly restrict discretion include:

Restricting the Second Company’s ability to issue securities senior to securities owned by the First Company
Requiring the Second Company to provide reports or other information of the type ordinarily available to common stockholders
Requiring that the First Company be able to purchase additional securities issued by the Second Company in order to maintain its percentage ownership in the Second Company
Requiring that the Second Company ensure that any securityholder that intends to sell its securities of the Second Company provide other securityholders of the Second Company or the Second Company the opportunity to purchase the securities first
Requiring the company take reasonable steps to ensure the preservation of tax status or tax benefits

III. Federal Reserve Math for Calculating Percentage of Voting Shares

The various presumptions in the Control Rule are keyed off percentages of classes of voting shares. The Control Rule largely retains existing practice for calculating these percentages, which is highly conservative: if an investor has options, warrants, or securities that are convertible at its option into voting shares of the Second Company, it controls the maximum number of voting shares that it could hold if it exercised all its options and all similar rights held by other investors are not exercised at all. There are no exceptions for out-of-the-money options. One exception to this general rule is if, by the terms of the instrument, an investor’s rights may be exercised only simultaneously with other investors. Second, and consistent with existing practice, there is an exception for preferred stock and similar instruments that give a holder the right to elect directors if dividends are not paid for a period of time and that otherwise have no voting rights; they are deemed voting securities only when the right to vote arises. A third is a purchase agreement to acquire securities that has not closed. A fourth is for rights to maintain an investor’s current voting share percentage.

In addition, if an investor is subject to an agreement whereby the rights of another holder of voting shares is restricted by the investor, it is deemed to control those shares. The Control Rule provides six exceptions to this rule:

Market terms rights of first refusal; rights of last refusal; tag-along rights; and drag-along rights
Restrictions incidental to a bona fide loan
Lock-up restrictions pending the consummation of an acquisition
An arrangement that requires a current shareholder of a company to vote in favor of an acquisition
Restrictions necessary to preserve a particular tax status or tax assets against impairment
Short-term revocable proxies

In terms of percentage of votes, the Control Rule states that an investor controls the greater of (i) the number of voting shares it controls divided by the number of issued and outstanding voting shares of that class, and (ii) the number of votes that the investor could cast divided by the number of votes that may be cast under all the issued and outstanding voting shares of that class.

IV. Restrictions on Non-Voting Shares

The Control Rule retains the Federal Reserve’s historical limitations on the ability of shares that are initially non-voting to become voting. These limitations are strict, and therefore may affect the attractiveness of holding substantial blocs of non-voting shares notwithstanding the Control Rule’s permitted percentage of total equity of up to one-third.

Non-voting shares may become voting shares in the hands of a transferee only in the following circumstances:

In a transfer back to the issuer of the shares
In a widespread public distribution
In transfers in which no transferee or group of associated transferees would receive 2 percent or more of any class of outstanding voting securities of the issuer
In a transfer to a transferee that would control more than 50 percent of every class of voting securities of the issuer without counting the transferred shares.

Such shares are also subject to strict transfer limitations.

V. Federal Reserve Math for Calculating Total Equity

The Control Rule’s general approach to calculating the percentage of total equity owned by an investor is based on U.S. GAAP, with the Federal Reserve stating that if the Second Company is not a stock corporation or does not prepare U.S. GAAP financial statements, total equity will be calculated “so as to be reasonably consistent” with the U.S. GAAP methodology, “taking into account the legal form of the second company and [its] accounting system.”

Under the general approach, the first step is to determine the percentage of each class of voting and nonvoting common or preferred stock issued by the Second Company that the First Company controls. Pari passu preferred stock is treated as one class of preferred stock. One then multiplies the percentage of each class of stock controlled by the First Company by the value of shareholders’ equity allocated to the class of stock under GAAP, with retained earnings allocated to common stock. The final step is to divide the First Company’s value of shareholders’ equity, as calculated under the previous test, by the total value of the Second Company’s shareholders’ equity, as determined under GAAP. The percentage of total equity is calculated at the time of the First Company’s investment and is required to be recalculated only if and when the First Company acquires additional equity. Although this method of calculating total equity was criticized by commenters with justification, the Federal Reserve was unwilling to depart from the test in the Control Rule, which may hamper certain investments.

VI. Special Rule for Investment Funds

There is a special presumption of control for investment funds. A company will be presumed to control an investment fund if it is the fund’s investment adviser and directly or indirectly, or acting through one or more persons, controls 5 percent or more of any class of voting securities of the fund, or controls 25 percent or more of the total equity of the fund. To allow for a seeding period, the presumption does not apply if the investor organized and sponsored the investment fund within the preceding 12 months.

VII. Divestiture of Control

The Control Rule modifies the Federal Reserve’s traditional “tear down” rules under which an investor that had BHC control of a company was required to divest its shares below 10 percent voting share ownership in order to shed control. The Control Rule states that a sale to below 15 percent will divest control once two years have passed. There is an exception if after the divestment, 50 percent or more of each class of voting securities of the divested company is controlled by a person that is not a senior management official or director of the divesting company, or by a company that is not an affiliate of the divesting company.

Conclusion

The Control Rule is a victory for regulatory transparency in that after over 40 years of interpretations, the Federal Reserve has used the notice and comment process to promulgate a regulation on this most important aspect of bank holding company law. Investors – both those who wish to invest in banking organizations, and banking organizations that wish to make non-controlling investments themselves – have gained some incremental benefits, although these are really focused at the level of investments of 9.99% voting shares and below. The framework in the Control Rule, moreover, is just that – a framework. Particular investments will continue to have particular facts that the framework will not clearly answer and for which judgment will be required.

APPENDIX 1

Summary of Tiered Presumptions

(Presumption triggered if any relationship exceeds the amount on the table)

	Less than 5% voting	5-9.99% voting	10-14.99% voting	15-24.99% voting
Directors	Less than half	Less than a quarter	Less than a quarter	Less than a quarter
Director Service as Board Chair	N/A	N/A	N/A	No director representative is chair of the board
Director Service on Board Committees	N/A	N/A	A quarter or less of a committee with power to bind the company	A quarter or less of a committee with power to bind the company
Business Relationships	N/A	Less than 10% of revenues or expenses of the second company	Less than 5% of revenues or expenses of the second company	Less than 2% of revenues or expenses of the second company
Business Terms	N/A	N/A	Market Terms	Market Terms
Officer/Employee Interlocks	N/A	No more than 1 interlock, never CEO	No more than 1 interlock, never CEO	No interlocks
Contractual Powers	No management agreements	No rights that significantly restrict discretion	No rights that significantly restrict discretion	No rights that significantly restrict discretion
Proxy Contests (directors)	N/A	N/A	No soliciting proxies to replace more than permitted number of directors	No soliciting proxies to replace more than permitted number of directors
Total Equity	BHCs – Less than 1/3 SLHCs – 25% or less	BHCs – Less than 1/3 SLHCs – 25% or less	BHCs – Less than 1/3 SLHCs – 25% or less	BHCs – Less than 1/3 SLHCs – 25% or less

______________________

[1] In the case of an investment in a savings-and-loan holding company, the relevant total equity percentage is 25% or less in all cases.

[2] In the case of non-GAAP financial statements, the Federal Reserve will rely on the non-GAAP financials, “while taking differences in accounting standards into account as appropriate.”

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work in the firm’s Financial Institutions Practice Group, or the following:

Financial Institutions Group:
Arthur S. Long – New York (+1 212-351-2426, [email protected])
Matthew L. Biben – New York (+1 212-351-6300, [email protected])
Michael D. Bopp – Washington, D.C. (+1 202-955-8256, [email protected])
Stephanie Brooker – Washington, D.C. (+1 202-887-3502, [email protected])
M. Kendall Day – Washington, D.C. (+1 202-955-8220, [email protected])
Michelle M. Kirschner – London (+44 20 7071 4212, [email protected])
Jeffrey L. Steiner – Washington, D.C. (+1 202-887-3632, [email protected])
James O. Springer – New York (+1 202-887-3516, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

London partner Ali Nikpay is the author of “Clampdown on mergers could make Britain less competitive” [PDF] published by The Telegraph on February 14, 2020.

The recent case of Roche/Spark[1], in which Gibson Dunn acted for Spark, shows that the UK Competition and Markets Authority (the “CMA”) is willing to intervene in transactions on the basis of a company’s R&D efforts alone, i.e. in the absence of a commercialised product. This case is particularly relevant to pharmaceutical companies. It also has potentially wider implications for other markets where R&D is an integral part of the supply process.

In Roche/Spark, the CMA found that it had jurisdiction to review the USD 4.3 billion transaction based on the fact that Roche (the acquirer) was marketing products in the UK in the area of haemophilia A treatments and Spark (the target) was active in the clinical development of gene technology (“GT”) treatments for haemophilia A.

The basis for establishing jurisdiction was the ‘UK share of supply test’, which requires that both parties be active in an overlapping area of supply in the UK. The CMA for this purpose concluded that a firm engaged in R&D activities relating to haemophilia A treatments in the UK (in particular activities at a ‘relatively advanced staged’, meaning Phase II or more of clinical development) is active in the process of ‘supplying’ such pharmaceutical treatments in the UK.

It is also of interest that the vast majority of Spark’s activities were found to take place abroad and all of its ‘UK’ activities fell within the context of a global R&D programme. Regardless, the CMA still found sufficient UK nexus to review the transaction, based on limited factors, including future UK links.

This decision highlights the increasingly interventionist approach that the CMA is taking with respect to M&A activity in sectors where innovation is important – the CMA being particularly alert to the possibility of so called ‘killer acquisitions’ in these areas.

The approach adopted in this case is consistent with statements made by the CEO of the CMA previously, that the CMA will use the share of supply test flexibly in innovative markets in order to investigate the rationale behind a high target valuation: “In relation to jurisdiction, […] we can also consider the parties’ combined share of supply, and exercise jurisdiction if this exceeds 25% and any kind of increment in share is brought about by the deal. This is a flexible test which, in practice, has meant that the CMA has consistently been able to exert jurisdiction over transactions in digital markets, for example where the turnover of the target was limited, but the value of the deal was high” (emphasis added).

The same speech also signalled an interest in reviewing high-value deals for potential “killer acquisition” effects, noting that: “if the price paid by the acquirer seems hard to explain based on current or likely future earnings, we should scrutinise the rationale for the acquisition with particular rigour and consider, in particular, whether the purchase price could reflect the benefit of killing off emerging competition.”[2]

The Roche/Spark decision provides a practical demonstration of how the CMA may seek to establish jurisdiction in these cases.[3] It also provides a timely reminder that, going forwards, parties need to be wary of the jurisdictional approach of the CMA when transacting in innovative markets. The rationale and basis for establishing jurisdiction in this particular case is further described below. We have also outlined, at the end of this briefing, some considerations for current and future transactions based on this decision.

Jurisdictional Approach in Roche/Spark – the UK Share of Supply Test

The CMA’s decision to review the Roche/Spark transaction was based on a wide interpretation of the share of supply test and full use of its discretion when applying this test. In particular, the CMA found that the statutory definition of ‘supply’ is broad enough that it does not require actual sales.

Approach to the concept of ‘active in the supply’ of a pharmaceutical treatment

Spark was found to be ‘active in the supply of haemophilia A treatments in the UK’, despite the fact that its products were still in clinical development and have not yet been put on the market.

The CMA’s conclusion was based on an expansive interpretation of the competitive process. The CMA relied on the fact that the supply of a product in the pharmaceuticals sector encompasses several stages, with the R&D stage being an integral part of the process of supplying pharmaceutical treatments. The CMA then said that it found evidence showing that ‘significant competition’ exists between firms before their products are fully commercialized. That is, a firm with a currently-marketed product will alter its commercial strategy to compete against a product that is still in (albeit at a relatively developed stage of) clinical development and vice versa. This conclusion seems largely based on broad principles concerning the characteristics of pharmaceutical markets,[4] with only limited high-level examples given in the decision of the type of evidence actually relied on.[5] On this basis, the CMA found that its jurisdictional assessment could extend beyond a consideration of already marketed products, in order to reflect the commercial realities by which companies in this area interact.

A similar approach could be followed in future pharmaceuticals transactions, as well as in other sectors where R&D is an integral part of the process, particularly if there is some transparency as to competitors’ pipeline product development efforts and some evidence of competitive response.

UK nexus established based on minimal activities in the context of a global R&D programme

The CMA’s finding that Spark was active in the supply of GT haemophilia A treatments in the UK, was similarly based on a wide view of what is required for a UK nexus.

As the CMA accepted, most of Spark’s activities took place abroad: “[m]ost of Spark’s current Hem A R&D activities are carried out in Philadelphia, where Spark’s laboratory space and R&D professionals dedicated to SPK-8011 and SPK-8016 are located. Spark’s clinical trials for SPK-8011 and SPK-8016 are also managed from Spark’s Philadelphia-based operations. Of Spark’s eleven open clinical trial sites for SPK-8011, two are outside of the United States.”

Nonetheless, the CMA found a UK nexus based on the fact that R&D activities form an integral part of the process of making treatments available in the UK, and that several of Spark’s activities relevant to R&D and/or intended commercialisation took place in the UK – namely, Spark held certain UK/EU patents, had some employees undertaking activities in the UK and intended to include UK sites in clinical trials.

A UK nexus was thus established based on what can only be understood as minimal UK-based activities compared to Spark’s global R&D platform, together with future intentions as to (a) making treatments (if successfully developed) available in the UK and (b) conducting future clinical trials in the UK.

Frame of reference and measures for assessing the 25% threshold

In measuring whether the 25% share of supply test was met, the CMA looked at activities relating to the supply of ‘novel non-GT and GT haemophilia A treatments (including commercialised treatments and pipeline treatments which are at the Phase II or more advanced stage of clinical development)’ and used the specific metrics of (a) the number of full-time equivalent UK-based employees engaged in such activities; and (b) the ‘procurement’ of UK patents[6].

As a preliminary point, it is worth noting that the CMA’s focus on “novel” treatments, to the exclusion of the many traditional haemophilia A treatments, made it far more likely that the share of supply test would be met in this case.[7] Future cases involving innovative companies may also see a similarly narrow focus at the jurisdictional stage.

The second point to note is that the CMA relied on somewhat opaque and difficult to measure metrics in this case, which may prove a challenge for companies to replicate when considering whether to notify future deals.

In particular, in terms of the methodology adopted for calculating the shares, the CMA stated in its decision that a ‘simple quantitative counting exercise’ had been conducted (with no consideration of qualitative differences between employee or patent units). However, counting patents is well recognised to not be a straightforward exercise and the exact scope of employee activities taken into account was not explained.

Further, the calculations performed by the CMA in this case were based on data collected directly from various industry players. Transacting parties are likely to find it difficult to estimate internally an equivalent when their deal involves R&D. For example, information on the number of employees working on project pipelines or other R&D tasks are confidential. One cannot necessarily make inferences based on a merging party’s own set up, as there is no particular reason why a competitor would follow a similar approach in the UK in order to compete effectively in this area.

Finally, it is interesting that the metrics relied on by the CMA have only an indirect link to the ultimate substantive assessment. Metrics concerning the number of patents granted and straight counts of UK-based employees have little relation to the likely success and competitive impact of the ultimately-developed products or even to any qualitative assessment of the extent and strength of current R&D activities.

What does this decision mean for current and future transactions?

This decision signals very clearly that the CMA’s jurisdictional thresholds are viewed by the CMA as a technical barrier to overcome in cases where the CMA wishes to investigate the potential effects of a merger or acquisition.

For companies active in industries where competition exists between firms before their products are fully commercialised (stimulated by transparency in R&D activities) or where R&D is an integral part of the supply process, it is clear that the absence of target turnover in the UK or overlaps in directly marketed products or services will not be a bar to the CMA seeking to take jurisdiction.

Consequently, in such industries (and particularly in the pharmaceuticals sector) going forward, the possibility of UK CMA intervention must be considered even where one or both of the parties do not have UK turnover, if the parties’ actual and potential or potential products (as applicable) can be categorised within the same category of products. The risk of CMA intervention is likely to increase the further along the development process a pipeline product is and where there is evidence that the potential treatment/product is intended be marketed in the UK and/or activities are taking place in the UK with respect to the relevant product.

_________________

[1] ME/6831/19 – Anticipated acquisition by Roche Holdings, Inc. of Spark Therapeutics, Inc. See full decision here: https://assets.publishing.service.gov.uk/media/5e3d7c0240f0b6090c63abc8/2020207_-_Roche_Spark_-_non-confidential_Redacted-.pdf

[2] See the full version of Andrea Coscelli’s speech to the OECD/G7 conference on competition and the digital economy in June 2019 here: https://www.gov.uk/government/speeches/competition-in-the-digital-age-reflecting-on-digital-merger-investigations.

[3] Another recent example of the CMA taking steps to investigate a merger where one party has no market facing activities in the UK includes the CMA’s decision to investigate the acquisition by Takeaway.com, a leading online food delivery marketplace in Continental Europe, of Just Eat, one of the main players in the UK food delivery marketplace.

[4] Specifically, it states that many pharmaceuticals markets are characterized by: (i) significant switching costs (in particular, because patients may be highly reluctant to switch treatments) which it felt meant that, in practice, incumbent firms have an incentive to compete against known pipeline products so as to win as many patients as possible before such products come to market and reduce the eventual impact of such products on coming to market; and (ii) a degree of transparency meaning that pipeline product development can be tracked.

[5] The evidence of competitive interactions relied on (based on the un-redacted text) includes evidence showing that: (i) global marketing efforts, efforts to establish a strong presence in the haemophilia community and efforts to encourage stronger rates of uptake for marketed products are ‘in part’ a competitive reaction to pipeline products; (ii) third parties closely monitor the treatment pipeline and track the development of individual pipeline products (competitors, clinicians and haematologists); and (iii) companies active in R&D invest in promotion activities prior to products being commercialized, to ensure brand recognition before products come to market.

[6] Including EU patents validated in the UK.

[7] The CMA distinguished: (i) traditional non-GT from novel non-GT and GT haemophilia A treatments for this purpose; and (ii) pipeline treatments at Phase II (or more advanced) stage of clinical development from pipeline treatments at Phase I stage of clinical development (or earlier). Notably, the CMA went on to assess the impact of the merger in relation to the supply of all haemophilia A prophylactic treatments, including traditional FVIII treatments.

The following Gibson Dunn lawyers assisted in preparing this client update: Ali Nikpay, Deirdre Taylor and Sarah Parker, who represented Spark before the UK CMA. Adam di Vincenzo and Richard Parker secured clearance for the deal from the US Federal Trade Commission.

Gibson Dunn’s lawyers are available to assist in addressing any questions that you may have regarding the issues discussed in this update. For further information, please contact the Gibson Dunn Lawyer with whom you work, any member of the firm’s Antitrust & Competition practice group, or the authors:

Ali Nikpay – London (+44 (0)20 7071 4273, [email protected])
Deirdre Taylor – London (+44 (0)20 7071 4274, [email protected])
Sarah Parker – London (+44 (0)20 7071 4073, [email protected])

Please also feel free to contact any of the following practice leaders and members:

London
Patrick Doris (+44 20 7071 4276, [email protected])
Charles Falconer (+44 20 7071 4270, [email protected])
Ali Nikpay (+44 20 7071 4273, [email protected])
Philip Rocher (+44 20 7071 4202, [email protected])
Deirdre Taylor (+44 20 7071 4274, [email protected])

Brussels
Peter Alexiadis (+32 2 554 7200, [email protected])
Attila Borsos (+32 2 554 72 11, [email protected])
Jens-Olrik Murach (+32 2 554 7240, [email protected])
Christian Riis-Madsen (+32 2 554 72 05, [email protected])
Lena Sandberg (+32 2 554 72 60, [email protected])
David Wood (+32 2 554 7210, [email protected])

Munich
Michael Walther (+49 89 189 33 180, [email protected])
Kai Gesing (+49 89 189 33 180, [email protected])

Hong Kong
Kelly Austin (+852 2214 3788, [email protected])
Sébastien Evrard (+852 2214 3798, [email protected])

Washington, D.C.
D. Jarrett Arp (+1 202-955-8678, [email protected])
Adam Di Vincenzo (+1 202-887-3704, [email protected])
Scott D. Hammond (+1 202-887-3684, [email protected])
Kristen C. Limarzi (+1 202-887-3518, [email protected])
Joshua Lipton (+1 202-955-8226, [email protected])
Richard G. Parker (+1 202-955-8503, [email protected])
Cynthia Richman (+1 202-955-8234, [email protected])
Jeremy Robison (+1 202-955-8518, [email protected])
Brian K. Ryoo (+1 202-887-3746, [email protected])
Chris Wilson (+1 202-955-8520, [email protected])

New York
Eric J. Stock (+1 212-351-2301, [email protected])

Los Angeles
Daniel G. Swanson (+1 213-229-7430, [email protected])
Samuel G. Liversidge (+1 213-229-7420, [email protected])
Jay P. Srinivasan (+1 213-229-7296, [email protected])
Rod J. Stone (+1 213-229-7256, [email protected])

San Francisco
Rachel S. Brass (+1 415-393-8293, [email protected])

Dallas
Veronica S. Lewis (+1 214-698-3320, [email protected])
Mike Raiff (+1 214-698-3350, [email protected])
Brian Robison (+1 214-698-3370, [email protected])
Robert C. Walters (+1 214-698-3114, [email protected])

Denver
Ryan T. Bergsieker (+1 303-298-5774, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

On 30 January 2020, the Director-General of the World Health Organization declared that the outbreak of novel coronavirus (2019-nCoV) met the criteria for a Public Health Emergency of International Concern (“PHEIC”).[1] In order to qualify as a PHEIC, an event must: (i) be extraordinary; (ii) constitute a public health risk to other States through the international spread of disease; and (iii) potentially require a coordinated international response.[2] Since the PHEIC regulations were enacted 15 years ago, a PHEIC has been declared only on five previous occasions: swine flu (2009); polio (2014); zika (2016); and ebola (2014 and 2019).[3]

Coronavirus is taking a tremendous human toll and our thoughts go out to those affected. Recent reports also suggest that coronavirus is beginning to impact commercial operations as facilities are shut down or production is curtailed, workforces are sent home, and demand falls. In particular, recent press coverage suggests that this issue has arisen in the LNG market, with Chinese buyers citing the impact of coronavirus in connection with the failure to take delivery of LNG.[4]

Coverage also suggests that there are concerns about a tightening of demand and the potential effect of coronavirus on the supply chain for other critical commodities such as coal and iron ore.[5]

As the effects of coronavirus continue to impact the markets, contractual counterparties will look to the terms of their contracts where they are unable to perform. One possibility that has already been reported in the LNG markets (and is being discussed in the iron ore markets) is a declaration of force majeure.

This client alert examines common formulations used in force majeure provisions in LNG sale and purchase agreements (“SPAs”), considers the scope of force majeure clauses under English and New York law, and identifies points that clients should consider when negotiating and seeking to enforce force majeure clauses.

Force Majeure Clauses in LNG Sale and Purchase Agreements

The concept of a communicable disease crisis (we address the distinction between “epidemics” and “pandemics” below) is sometimes included in force majeure clauses in LNG SPAs and has been seen in contracts for delivery into various global markets, including Europe and Asia. Where included, however, it is typically still set out as part of a non-exhaustive list of events which may constitute force majeure provided that the general requirements for force majeure are also met. These usually require performance to have been prevented, hindered (or impaired), or delayed due to an event not within the reasonable control of the party seeking to claim force majeure, acting as a reasonable and prudent operator. They may also require that the event be unforeseen and that the claiming party has endeavoured to overcome or to mitigate the force majeure event, using reasonable diligence. In addition, some LNG SPAs require the event to have affected the relevant facilities in a specific way, such that the facilities are damaged or not capable of being operated due to the relevant event, in order for force majeure to be available as a remedy. Finally, since a force majeure clause is typically drafted as a general test with (more often than not) a non-exhaustive list of examples, the LNG SPA will contain a long list of excluded events (such as financial hardship, lack of demand or economic downturns). These types of qualifiers and exclusions mean that the occurrence of an epidemic, on its own, may not trigger force majeure relief.

References to “quarantine restrictions” are also sometimes included in non-exhaustive lists of force majeure events. Whilst we expect such clauses were likely intended to cover the quarantine of vessels, it remains to be seen whether this may also be relied upon where a party is unable to operate a facility due to quarantine restrictions affecting its workforce.

Under LNG SPAs there may also be limitations on the period or the quantities of LNG in respect of which an affected facility can continue to claim force majeure relief where the contract contains optionality as to delivery locations.

There is considerable divergence among LNG SPAs on the treatment of volumes which have been excused for force majeure. Some LNG SPAs require volumes to be taken at a later date (or require a buyer to use reasonable endeavours to take volumes at a later date) in order to restore the overall quantity delivered to the contracted quantity. Typically, these restoration quantities would be taken at the market price at the time of actual delivery, which may be either higher or lower than the time of the force majeure event. Naturally, this can create winners and losers depending on market fluctuations and may lead to assertions of arbitrage against the party claiming force majeure. Capacity and scheduling issues will also need to be addressed.

Other LNG SPAs simply remove the volume from the contract quantity, leaving the seller with additional volumes to market. Whilst this may be attractive for both sellers and buyers during periods where prices are high and cargoes can be diverted, it is likely to be less attractive in the current environment of oversupply and low prices.

Extended force majeure must also be considered in the context of an ongoing health crisis. Termination for extended force majeure appears as a remedy in many LNG SPAs (although where the LNG SPA underpins the financing of a liquefaction facility, financiers may have pushed back on this provision). The inclusion of a termination right may actually prevent a seller from triggering a force majeure claim if the continuation of the contract is more valuable to it where it is fixed with high prices (notwithstanding the possible force majeure event).

However, neither the presence of remedies nor the inclusion of a reference to epidemic or quarantine, of itself, guarantees relief. Instead, these provisions must be read and interpreted with the rest of the force majeure clause and the remainder of the contract. We consider this further below.

The Position Under English Law

“Force majeure” is not a term of art under English law. Accordingly, a force majeure clause in a commercial contract will be construed in each case on its own terms.[6] As the English Court of Appeal explained recently:

“[T]he question is one of construction … and the answer to that question is determined by the language of the clause which the parties have chosen, having regard to the context and purpose of the clause.”[7]

The usual rules of contractual interpretation under English law apply to the interpretation of force majeure clauses. The court must ascertain the objective meaning of the language which the parties have chosen to express their agreement, and consider not only the wording of the particular force majeure clause but must also consider the contract as a whole and, depending on the nature, formality and quality of drafting of the contract, give more or less weight to elements of the wider context in reaching its view as to that objective meaning.[8] Market practice may also be relevant to the exercise of interpretation, provided that it is clearly evidenced.[9] In practice, however, market practice may be difficult to prove and a recent English case has confirmed that evidence of market practice after a contract is concluded will not result in terms being implied into a contract.[10]

Force majeure clauses typically cover events that “prevent” a party from performing its contractual obligations or that “hinder” or “delay” performance. Therefore, when invoking force majeure clauses in response to events that prevent performance, it is usually necessary to establish that performance has become legally or physically impossible. Mere difficulty or unprofitability generally will not suffice.[11] By contrast, clauses that refer to a party being hindered or delayed have been construed more widely, although a change in economic circumstances or a market downturn affecting profitability or the ease with which obligations can be performed is still unlikely to qualify as a force majeure event in the eyes of the law,[12] and, as noted above, is commonly expressly carved out of force majeure relief in a well-drafted contract to avoid misuse of the clause. In either case, causation is key. In a recent case, the English High Court held that where there are multiple potential causes, some falling within a force majeure clause and some without, the force majeure must be the sole cause of the failure to perform the obligation.[13]

The Position Under New York Law

New York law on the interpretation of force majeure clauses generally aligns with English law. As with all contractual disputes, the terms of the agreement and the intent of the parties control a court’s analysis. Force majeure clauses will only limit damages where circumstances beyond the parties’ control have frustrated the parties’ “reasonable expectation[s].”[14] Force majeure clauses are construed narrowly, and “the general words are not to be given expansive meaning; they are confined to things of the same kind or nature as the particular matters mentioned.”[15] Moreover, some courts have held that the event must not only be included in the force majeure clause; it “must be unforeseeable as well,” although others have refused to impose an unforeseeability requirement where one cannot be found in the parties’ agreement.[16]

Notwithstanding these principles, New York courts will take account of well-established business practices in order to determine the parties’ intent, particularly in highly technical industries such as oil and gas. In a recent decision on questions certified by the Second Circuit Court of Appeals, the New York Court of Appeals confirmed that, consistent with the practices of other “oil jurisdictions” such as Texas and California, force majeure clauses “must be construed with reference to both the intention of the parties and the known practices within the industry.”[17]

Accordingly, the terms of the relevant contract are the starting point in any analysis of a force majeure event. But industry custom is important too, particularly in contracts relating to well-developed, technical industries such as oil and gas, where the parties’ intent may be gleaned from established practices. In all cases, however, force majeure clauses are construed narrowly and their terms will not be expanded beyond events of the same kind or nature as those listed in the parties’ agreement.

Drafting Tips

In light of the increased risk of global health crises, it is likely that this issue will continue to be widely and explicitly addressed in force majeure clauses within commodities contracts. As contractual language is the predominant factor in interpreting force majeure clauses, contracting parties should carefully consider how they draft such provisions to allocate risk appropriately.

Notwithstanding the increased frequency of disease outbreaks, absent express contractual protection, courts are likely to continue to construe force majeure clauses narrowly. We pause to note that parties should be mindful of the distinction between epidemics, which are local or regional, and pandemics, which are epidemics with global reach. When drafting force majeure clauses, the parties should consider the point at which the outbreak of a communicable disease would sufficiently impact their business relationship so as to trigger a force majeure clause.

When drafting force majeure clauses to account for disease outbreaks, the following complicating factors should be considered: (i) the difficulty of distinction between natural and political events; (ii) whether the event has a direct or indirect effect on performance, and the extent to which it affects performance; (iii) inability to perform versus desire to escape performance due to unprofitability; (iv) the types of mitigating actions that an affected party should undertake and (v) the possibility of an extended force majeure event. First, force majeure clauses often distinguish between natural and political events.[18] Although of natural origin, the economic impact of an epidemic will flow in large part from governmentally imposed quarantines and travel restrictions. To avoid claims that a natural event force majeure clause has not been triggered because the prevention or hindering of performance is the result of government intervention, contracts should be clear on any distinction between natural and political events, or make express provision for government-imposed measures responding to the outbreak of infectious disease. Second, parties should consider the real extent to which performance needs to be prevented by the event claimed as force majeure or the extent to which there could be a combination of events that prevent performance. Where a facility is unable to operate due to workforce unavailability, consider the extent to which it must be shown that this is a result of the force majeure event. Third, force majeure clauses typically contain language on the exercise of reasonable diligence or efforts to overcome an event. Rather than rely on case law to interpret what these measures might or might not involve, specific examples could be included. Fourth, as stated, a typical force majeure clause will usually only offer protection where performance is genuinely hindered, prevented or delayed. Where an event leads to reduced demand (and potentially reduced prices in a spot market), the stakes can be raised for a variable-price seller or a fixed-price buyer. Force majeure clauses should be crafted carefully or such risks should be otherwise addressed. Fifth, epidemics pose a relatively high risk of creating an extended force majeure event. Contracting parties should consider whether termination rights should be included, and whether to set a minimum duration for a force majeure event before a party may elect to enforce any termination rights.

_____________________

[1] World Health Organization, Statement on the second meeting of the International Health Regulations (2005) Emergency Committee regarding the outbreak of novel coronavirus (2019-nCoV) (Jan. 30, 2020), available at https://www.who.int/news-room/detail/30-01-2020-statement-on-the-second-meeting-of-the-international-health-regulations-(2005)-emergency-committee-regarding-the-outbreak-of-novel-coronavirus-(2019-ncov).

[2] See International Health Regulations, art. 1(1) (2005).

[3] See James Gallagher, Coronavirus declared global health emergency by WHO, BBC News (Jan. 31, 2020), available at https://www.bbc.co.uk/news/world-51318246.

[4] See Stephen Stapczynski, China LNG Force Majeure Rejected as Virus Chaos Sparks Dispute, Bloomberg News (Feb. 7, 2020), available at https://www.bloomberg.com/news/articles/2020-02-07/china-lng-force-majeure-rejected-as-virus-chaos-sparks-dispute; Alaric Nightingale & Alex Longley, When God Appears in Contracts, That’s ‘Force Majeure’, Bloomberg News (Feb. 6, 2020), available at https://www.bloomberg.com/news/articles/2020-02-06/when-god-appears-in-contracts-that-s-force-majeure-quicktake.

[5] See Deepali Sharma, Coronavirus impact on iron ore market, American Metal Market (Feb. 6, 2020), available at https://www.amm.com/Article/3917527/Coronavirus-impact-on-iron-ore-market.html; Frick Els, Coronavirus drops iron ore shipping gauge 99.9%, Mining.com (Jan 31, 2020), available at https://www.mining.com/coronavirus-drops-iron-ore-shipping-gauge-99-9/; Jun Kai Heng, Niki Wang, & Yuchen Huo, Iron ore outlook bearish as steel demand seen weak on coronavirus, S&P Global (Feb. 5, 2020), available at https://www.spglobal.com/platts/en/market-insights/latest-news/metals/020520-iron-ore-outlook-bearish-as-steel-demand-seen-weak-on-coronavirus.

[6] The English courts will not imply a force majeure clause or, where there is such a clause, imply particular events within it. However, in the absence of an express contractual provision catering for the consequences of the event that has occurred, it may still be possible for a party facing impossibility of performance to rely on the common law doctrine of frustration. The operation of the doctrine of frustration requires an unforeseen, supervening event which renders it physically or commercially impossible for a party to fulfil a contract or transforms the obligation to perform into a radically different obligation from that undertaken at the moment of entry into the contract and operates to bring a contract to an end “forthwith, without more and automatically.” See Chitty on Contracts (33rd ed.), §§ 23-001, 23-007; J. Lauritzen AS v Wijsmuller BV (The Super Servant Two) [1990] 1 Lloyd’s Rep 1, 8.

[7] Classic Maritime Inc v Limbungan Makmur Sdn Bhd and another [2019] EWCA Civ 1102 at [32].

[8] Wood v Capita Insurance Services Limited [2017] UKSC 24 at [10].

[9] Crema v Cenkos Securities Plc [2010] EWCA Civ 1444.

[10] TAQA Bratani Ltd & Others v RockRose UKCS8 LLC [2020] EWHC 58 (Comm).

[11] Chitty on Contracts (33rd ed.), § 15-156.

[12] Chitty on Contracts (33rd ed.), §§ 15-158 to 15-160; Tandrin Aviation Holdings Limited v Aero Toy Store LLC and another [2010] EWHC 40 (Comm) at [40].

[13] Seadrill Ghana Operations Ltd v Tullow Ghana Ltd [2018] EWHC 1640 (Comm).

[14] See Constellation Energy Servs. of New York, Inc. v. New Water St. Corp., 146 A.D.3d 557, 558, (1st Dep’t 2017) (internal citation omitted) (emphasis added).

[15] Kel Kim Corp. v. Cent. Markets, Inc., 519 N.E.2d 295, 297 (N.Y. 1987) (citing 18 Williston, Contracts § 1968 (3d ed. 1978)).

[16] See Phibro Energy, Inc. v. Empresa De Polimeros De Sines Sarl, 720 F. Supp. 312, 318 n.8 (S.D.N.Y. 1989) (citing United States v. Brooks-Callaway Co., 318 U.S. 120, 122–23 (1943)); see also Goldstein v. Orensanz Events LLC, 146 A.D.3d 492, 493 (1st Dep’t 2017) (“[T]he clause must be interpreted as if it included an express requirement of unforeseeability or lack of control.”).

[17] Beardslee v. Inflection Energy. LLC, 31 N.E.3d 80, 84 (N.Y. 2015).

[18] See, e.g., World Bank, Force Majeure Clauses – Checklist and Sample Wording 8–9, available at http://siteresources.worldbank.org/INTINFANDLAW/Resources/Forcemajeurechecklist.pdf (last visited Feb. 11, 2020).

The following Gibson Dunn lawyers assisted in the preparation of this client update: Cyrus Benson, Anna Howell, Brad Roach, Zachary Kady, Mitasha Chandok, Nick Kendrick, Piers Plumptre, Philip Shapiro, Moeiz Farhan and Brandon Davis.

Gibson, Dunn & Crutcher’s lawyers are available to assist in addressing any questions you may have regarding these issues. Please contact the Gibson Dunn lawyer with whom you usually work, any member of the firm’s International Arbitration or Oil and Gas practice groups, the authors or any of the following:

Oil and Gas Group:
Anna P. Howell – London (+44 (0)20 7071 4241, [email protected])
Brad Roach – Singapore (+65 6507.3685, [email protected])
Saptak Santra – Singapore (+65 6507.3691, [email protected])
Mike Raiff – Dallas (+1 214-698-3350, [email protected])
Doug Rayburn – Dallas (+1 214-698-3442, [email protected])
Michael P. Darden – Houston (+1 346-718-6789, [email protected])
Tull Florey – Houston (+1 346-718-6767, [email protected])
Hillary H. Holmes – Houston (+1 346-718-6602, [email protected])
Shalla Prichard – Houston (+1 346-718-6644, [email protected])
Gerry Spedale – Houston (+1 346-718-6888, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

German criminal law so far does not provide for corporate criminal liability. Corporations can only be fined under the law on administrative offenses, and the government has discretion to impose such fines.

In August 2019, the German Federal Ministry of Justice and Consumer Protection circulated a legislative draft of the Corporate Sanctions Act which would introduce a hybrid system. The main changes to the current legal situation would eliminate the prosecutorial discretion in initiating proceedings, tighten the sentencing framework and formally incentivize the implementation of compliance measures and internal investigations.

Join our team of white collar experts to find out what companies will have to expect in the future when it comes to corporate crimes.

Topics covered:

Farewell to discretion – Mandatory prosecution of corporate crimes
New sentencing framework – Imposing harsher sanctions
Legal incentives for compliance, internal investigations and cooperation
Statutory requirements for internal investigations
Protecting documents against seizure

View Slides (PDF)

PANELISTS:

Finn Zeidler is a partner in the Frankfurt office and a member of the firm’s White Collar Defense and Investigations and Litigation Groups. Mr. Zeidler focuses his litigation practice on white collar defense, regulatory investigations and compliance matters that often have cross-border elements, as well as corporate and commercial litigation and arbitration, with an emphasis on securities and post-M&A disputes. He has significant experience in the banking and finance industry.

Ralf van Ermingen-Marbach is of counsel in the Munich office and a member of the firm’s White Collar Defense and Investigations Practice. Mr. van Ermingen-Marbach focuses his practice on internal investigations and the representation of corporations and individuals facing charges of criminal conduct. He has a decade-long experience as a prosecutor and has profound expertise in representing companies from the technology, banking, energy and health care industries, as well as funds and high-profile individuals.

In 2019, companies and regulators faced unprecedented challenges as they navigated a rapidly evolving set of issues and policy proposals on the regulation of Artificial Intelligence and Automated Systems (“AI”). Lawmakers grappled with the difficult questions of how and when to regulate AI and sketched out new legal frameworks, bringing into sharper relief the overarching legal issues that are poised to become the subject of protracted debate over the coming year. The policy debate in 2019 was especially characterized by the gulf between sprawling treatises setting out general ethical principles designed to control and mitigate the risks of AI—including theoretical applications of “general” AI[1]—on the one hand, and calls for targeted restrictions on the specific use of certain “narrow” domain-specific AI products and applications on the other.

In the United States, while federal policy remains “light-touch,” lawmakers responded to increasing public concern over the perceived dangers of unfettered technological development by proposing a number of high profile draft bills addressing the role of AI and how it should be governed, the real impact of which is yet to be felt across the private sector. U.S. state and local governments pressed forward with concrete legislative proposals regulating the use of AI. Finally, 2019 saw a growing international consensus that AI technology should be subject to certain technical standards and potentially even certification procedures in the same way other technical systems require certification before deployment.

This inaugural Artificial Intelligence and Automated Systems Annual Legal Review places these, and other, 2019 developments in broader context, focusing on developments within the United States. We also touch, albeit non-exhaustively, on developments within the EU that are of relevance to domestic and international companies alike. The AI policy landscape is vast and fast-evolving—we do not attempt to address every development that occurred in 2019, but examine a number of the most significant developments affecting companies as they navigate the evolving AI landscape.

__________________________

I. GLOBAL POLICY DEVELOPMENTS

A.  OECD Recommendations on AI
B.  Global Partnership on AI
C.  Global Initiative on Ethics of Autonomous and Intelligent Systems
D.  UN Considers Ban on Lethal Autonomous Weapons

II. U.S. NATIONAL POLICY & KEY LEGISLATIVE EFFORTS

A. U.S. National Policy on AI Takes Shape
B. U.S. Imposes Export Controls Related to AI

III. EU POLICY & REGULATORY DEVELOPMENTS

A.  EC Focus on Comprehensive AI Regulation
B.  Ethics Guidelines for Trustworthy AI
C.  German Data Ethics Commission Report

IV. REGULATION OF AI TECHNOLOGIES AND ALGORITHMS

A.  Algorithmic Accountability
B.  Facial Recognition Software
C.  Deepfake Technology
D.  Autonomous Vehicles
E.  Data Privacy
F.  Intellectual Property
G.  Law Enforcement
H.  Health Care
I.   Financial Services
J.  Labor and Hiring

__________________________

I. GLOBAL POLICY DEVELOPMENTS

The past year saw a number of ambitious projects and guidelines from governments and other intergovernmental bodies around the world aiming to identify shared norms and values—described by one such initiative as “holistic definitions of societal prosperity […] versus […] one-dimensional goals of increased productivity or gross domestic product”[2]—in an attempt to reach a global consensus on how to define and then instill an ethical approach into AI development and governance.[3] While these recommendations are non-binding and advisory in nature, their potential impact on national policy-making and, ultimately, concrete regulations, should not be underestimated. We offer a brief overview of several of these policy initiatives in order to place legislative, regulatory and policy approaches proposed and enacted by governments and other domestic regulatory bodies in a broader global context.

A. OECD Recommendations on AI

On May 21, 2019, the 36 member countries of the Organization for Economic Co-operation and Development (“OECD”), along with Argentina, Brazil, Colombia, Costa Rica, Peru and Romania, adopted a set of recommendations on AI referred to as “Principles of Artificial Intelligence” (“Principles”).[4] The Principles were drafted by a group of experts comprising experts from different OECD members, disciplines and sectors.[5] Intended to represent the first intergovernmental standard for AI policies, they promote five strategies for developing national policy and international cooperation, treading the line between promoting economic improvement and innovation and fostering fundamental values and trust in the development of AI: inclusive growth, sustainable development and well-being; human-centered values and fairness; transparency and explainability; robustness, security and safety; and accountability. The Principles are broadly stated and not legally binding, but instead seek to encourage member countries to incorporate these values or ethics in the development of AI.[6]

B. Global Partnership on AI

In May 2019, Canada and France announced plans for a new international body (“Global Partnership on AI”) for the G7 countries to study and steer the effects of artificial intelligence on the world’s people and economies by creating best practices, modeled on the UN’s Intergovernmental Panel on Climate Change.[7] Although the principles espoused by the Global Partnership would not be legally binding on other governments or private companies, the White House has raised concerns that the approach is too restrictive and duplicates work already being done by the OECD, and has so far declined to participate.[8]

C. Global Initiative on Ethics of Autonomous and Intelligent Systems

In 2019, the Institute of Electrical and Electronics Engineers (“IEEE”) launched the “Global Initiative on Ethics of Autonomous and Intelligent Systems” in order to “establish societal and policy guidelines in order for such systems to remain human-centric, serving humanity’s values and ethical principles.”[9] In an extensive report addressing ethical considerations in the design of AI systems, IEEE applied classical ethics and human rights methodologies to considerations of algorithmic design and articulated eight high-level ethical principles that apply to all types of AI products, regardless of whether they are physical robots or software systems, and that “define imperatives for the design, development, deployment, adoption, and decommissioning of [AI].”[10] The eight principles are: human rights, human well-being, data agency and control, effectiveness and fitness for purpose, transparency, accountability, awareness of misuse, and operational competence.[11]

D. UN Considers Ban on Lethal Autonomous Weapons

In March 2019, the United Nations (“UN”) Secretary-General António Guterres urged restriction on the development of lethal autonomous weapons systems (“LAWs”), arguing that machines with the power and discretion to take lives without human involvement are politically unacceptable, morally repugnant and should be prohibited by international law.[12] Subsequently, Japan pledged that it will not develop fully automated weapons systems.[13] A group of member states—including the UK, United States, Russia, Israel and Australia—are reportedly opposed to a preemptive ban in the absence of any international agreement on the characteristics of autonomous weapons, stalling any progress towards a common approach for the time being.[14]

II. U.S. NATIONAL POLICY & KEY LEGISLATIVE EFFORTS

A. U.S. National Policy on AI Takes Shape

By early 2019, despite its position at the forefront of commercial AI innovation, the United States still lacked an overall federal AI strategy and policy.[15] Under increasing pressure from the U.S. technology industry and policy organizations to present a substantive federal AI strategy on AI, over the past 12 months the Trump administration took public actions to prioritize AI and automated systems. Most notably, these pronouncements include President Trump’s “Maintaining American Leadership in Artificial Intelligence” Executive Order[16] and the subsequently issued guidance to federal regulatory agencies.[17] U.S. federal, state and local government agencies also began to show a willingness to take concrete positions on regulation, resulting in a variety of policy approaches – many of which eschew informal guidance and voluntary standards and favor outright technology bans. For the most part, the trend in favor of more individual and nuanced assessments of how best to regulate AI systems specific to their end uses by regulators in the United States has been welcome. Although there is an inherent risk that reactionary legislative responses will result in a disharmonious, fragmented national regulatory framework, such developments will yield important insights into what it means to govern and regulate AI over the coming year.

1. Executive Order “Maintaining American Leadership in Artificial Intelligence”

On February 11, 2019, President Trump signed an executive order (“EO”) titled “Maintaining American Leadership in Artificial Intelligence.”[18] The purpose of the EO was to spur the development and regulation of artificial intelligence, machine learning and deep learning and fortify the United States’ global position by directing federal agencies to prioritize investments in AI,[19] interpreted by many observers to be a response to China’s efforts to claim a leadership position in AI research and development.[20]

The EO, which was titled ‘Maintaining American Leadership in Artificial Intelligence,’ outlined five key areas: research and development,[21] ‘unleashing’ AI resources,[22] establishing AI governance standards, building an AI workforce,[23] and international collaboration and protection.[24] The AI Initiative is coordinated through the National Science and Technology Council Select Committee on Artificial Intelligence (“NSTC Select Committee”).

While the EO favors broad principles in line with the administration’s “light-touch” approach to private sector regulation, AI developers will need to pay close attention to the executive branch’s response to standards setting. Aiming to foster public trust in AI by using federal agencies to develop and maintain approaches for safe and trustworthy creation and adoption of new AI technologies (for example, the EO calls on the National Institute of Standards and Technology (“NIST”) to lead the development of appropriate technical standards).[25]

In response, in July 2019 NIST sought public comment on a draft plan for federal government engagement in advancing AI standards for U.S. economic and national security needs (“U.S. Leadership in AI: Plan for Federal Engagement in Developing Technical Standards and Related Tools”).[26] The plan recommends four actions: bolster AI standards-related knowledge, leadership and coordination among federal agencies; promote focused research on the “trustworthiness” of AI; support and expand public-private partnerships; and engage with international parties.

The full impact of the AI Initiative is not yet known: while it sets some specific deadlines for formalizing plans by agencies under the direction of the Select Committee, the EO is not self-executing and is generally thin on details. Therefore, the long-term impact will be in the actions recommended and taken as a result of those consultations and reports, not the EO itself.[27] Moreover, although the AI Initiative is designed to dedicate resources and funnel investments into AI research, the EO does not set aside specific financial resources or provide details on how available resources will be structured.

In March 2019, the White House launched ai.gov as a platform to share AI initiatives from the Trump administration and federal agencies.[28] These initiatives track along the key points of the AI EO, and ai.gov is intended to function as an ongoing press release, highlighting a number of federal government efforts under the Trump administration (and some launched during the Obama administration): the White House’s charting of the NSTC Select Committee on AI, the Department of Energy’s efforts to develop supercomputers, the Department of Transportation’s efforts to integrate automated driving systems, and the Food and Drug Administration’s efforts to assess AI implementation in medical research.[29]

2. The GrAITR Act (H.R. 2202)

The Growing Artificial Intelligence Through Research (GrAITR) Act was introduced in April 2019 to establish a coordinated federal initiative aimed at accelerating AI research and development for US economic and national security and closing the existing funding gap.[30] The Act would create a strategic plan to invest $1.6 billion over 10 years in research, development and application of AI across the private sector, academia and government agencies, including NIST, and the National Science Foundation and the Department of Energy – aimed at helping the United States catch up to other countries, including the United Kingdom, who are “already cultivating workforces to create and use AI-enabled devices.” The bill has been referred to the House Committee on Science, Space, and Technology.[31]

3. Artificial Intelligence Initiative Act (S. 1558)

In May 2019, U.S. Senators Rob Portman (R-OH), Martin Heinrich (D-NM), and Brian Schatz (D-HI) proposed a companion bill to GrAITR, the Artificial Intelligence Initiative Act, which would attempt to create a national, overarching strategy ‘tailored to the US political economy’, for developing AI with a $2.2 billion federal investment over the next five years.[32] The Act would task branches of the federal government to use AI where possible in operation of its systems. Specifically, it includes the establishment of a national office to coordinate AI efforts across the federal system (National AI Coordination Office), requests that NIST establish ethical standards and identify metrics used to establish standards for evaluating AI algorithms and their effectiveness, as well as the quality of training data sets, and proposes that the National Science Foundation set educational goals for AI and STEM learning.[33] Moreover, the bill requires the Department of Energy to create an AI research program, building state-of-the-art computing facilities that will be made available to private sector users on a cost-recovery basis.[34] The draft legislation complements the formation of the bipartisan Senate AI Caucus in March 2019 to address transformative technology with implications spanning a number of fields including transportation, healthcare, agriculture, manufacturing and national security.

4. AI in Government Act (H.R. 2575/S. 3502)

House Bill 2575 and its corresponding bipartisan Senate Bill 3502 (the “AI in Government Act”)—which would task federal agencies with exploring the implementation of AI in their functions and establishing an “AI Center of Excellence,”—were first introduced in September 2018, and reintroduced in May 2019.[35] The center would be directed to “study economic, policy, legal, and ethical challenges and implications related to the use of artificial intelligence by the Federal Government” and “establish best practices for identifying, assessing, and mitigating any bias on the basis of any classification protected under Federal non-discrimination laws or other negative unintended consequence stemming from the use of artificial intelligence systems.”

One of the sponsors of the bill, Senator Brian Schatz (D-HI), stated that “[o]ur bill will bring agencies, industry, and others to the table to discuss government adoption of artificial intelligence and emerging technologies. We need a better understanding of the opportunities and challenges these technologies present for federal government use and this legislation would put us on the path to achieve that goal.”[36] Although the bill is aimed at improving the implementation of AI by the federal government, there are likely to be opportunities for industry stakeholders to participate in discussions surrounding best practices.[37]

5. OMB Guidance for Federal Regulatory Agencies

The EO directed the Office of Management and Budget (“OMB”) director, in coordination with the directors of the Office of Science and Technology Police, Domestic Policy Council, and National Economic Council, and in consultation with other relevant agencies and key stakeholders (as determined by OMB), to issue a memorandum to the heads of all agencies to “inform the development of regulatory and non regulatory approaches” to AI that “advance American innovation while upholding civil liberties, privacy, and American values” and consider ways to reduce barriers to the use of AI technologies in order to promote their innovative application. The White House Office of Science and Technology Policy further indicated in April 2019 that regulatory authority will be left to agencies to adjust to their sectors, but with high-level guidance from the OMB, as directed by the EO.[38]

In January 2020, the OMB published a draft memorandum featuring 10 “AI Principles” and outlining its proposed approach to regulatory guidance for the private sector which echoes the “light-touch” regulatory approach espoused by the 2019 EO, noting that promoting innovation and growth of AI is a “high priority” and that “fostering innovation and growth through forbearing from new regulations may be appropriate.”[39] The guidance directs federal agencies to “avoid regulatory or non-regulatory actions that needlessly hamper AI innovation and growth” and notes that in certain circumstances it may be appropriate to preempt “inconsistent, burdensome and duplicative State laws,” although it also cautions that agencies should consider forgoing regulatory action where a “uniform national standard for a specific aspect related to AI is not essential.”[40] As expected, the principles favor flexible regulatory frameworks that allow for rapid change and updates across sectors, rather than one-size-fits-all regulations, and urge European lawmakers to avoid heavy regulation frameworks. The guidance encourages federal agencies to provide opportunities for public comment in AI rulemaking.

The principles also address some of the concerns raised by commentators with regard to ethics and particularly unwanted bias, urging lawmakers to consider whether the technology will “introduce real-world bias that produces discriminatory outcomes” and advising agencies to pursue transparency by disclosing the use of AI technology and making sure that outcomes are sufficiently transparent to ensure that the algorithms comply with existing laws.

6. National Security and Military Use

In the last few years, the US federal government has been very active in coordinating cross-agency leadership and planning for bolstering continued research and development of artificial intelligence technologies for use by the government itself. Along these lines, a principle focus for a number of key legislative and executive actions was the growth and development of such technologies for national security and military uses.

As a result of the passing of the John S. McCain National Defense Authorization Act for 2019 (the 2019 NDAA),[41] the National Security Commission on Artificial Intelligence was established to study current advancements in artificial intelligence and machine learning, and their potential application to national security and military uses. In addition, in response to the 2019 NDAA, the Department of Defense (“DoD”) created the Joint Artificial Intelligence Center (“JAIC”) as a vehicle for developing and executing an overall AI strategy, and named its director to oversee the coordination of this strategy for the military.[42] While these actions clearly indicate an interest in ensuring that advanced technologies like AI also benefit the US military and intelligence communities, the limited availability of funding from Congress may hinder the ability of these newly formed entities to fully accomplish their stated goals.

The JAIC is becoming the key focal point for the DoD in executing its overall AI strategy. As set out in a February 2019 summary of AI strategy provided by the DoD,[43] the JAIC will work with the Defense Advanced Research Projects Agency (“DARPA”),[44] various DoD laboratories, and other entities within the DoD to not only identify and deliver AI-enabled capabilities for national defense, but also to establish ethical guidelines for the development and use of AI by the military.[45] However, JAIC’s efforts to be a leader in defining ethical uses of AI in military applications may prove challenging because one of the most hotly debated uses of AI is in connection with autonomous weaponry.[46] As this Review went to press, the White House released its 2021 budget request to Congress, which proposed a funding windfall for AI-related research and development, particularly in the military sector.[47]

On October 31, 2019, the Defense Innovation Board (“DIB”), an independent federal advisory committee to the Pentagon consisting of a group of science and technology experts—led by former Google CEO Eric Schmidt—proposed a new ethics framework consisting of five overarching ethical principles which tie the DoD existing laws of war and rules of engagement[48] into the use of AI.[49] The report is a high-level blueprint for military deployments of artificial intelligence and addresses some general shortcomings of AI technology.[50] The principles advocate for deliberate AI designs to counter unintended biases that could cause inadvertent harm and for humans to have the power to deactivate or disengage AI systems acting outside the intended parameters. The DIB also suggested that humans should always be responsible for the “development, deployment, use and outcomes” of AI rather than letting AI set its own standards of use. In these cases, DoD should not use that AI system because “it does not achieve mission objectives in an ethical or responsible manner.”[51]

The DIB also recommended a number of technical and organizational measures that would help lay the groundwork to ensure military artificial intelligence systems adhere to ethical standards, such as increasing investment in standards development, workforce programs and AI security applications, and formalizing channels for exploring the ethical implications of deploying AI technology across the department. The newly proposed ethics framework could help address private sector concerns about innovative technology being wrongly weaponized or misused by the military or being part of autonomous systems without sufficient human oversight.

a) The National Artificial Intelligence Research and Development Strategic Plan: 2019 Update

Three years after the release of the initial National Artificial Intelligence Research and Development Strategic Plan, in June 2019 the Trump administration issued an update—previewed in the EO—bringing forward the original seven focus areas and adding an eighth: public-private partnerships.[52] The update highlights the benefits of strategically leveraging academic and industry expertise, including facilities, datasets, and expertise, to advance science and engineering innovations. Companies interested in exploring the possibility of individual collaborations or joint programs advancing precompetitive research should consider whether they have relevant expertise in any of the areas in which federal agencies are actively pursuing public-private partnerships, including the DoD’s Defense Innovation Unit and the Department of Health and Human Services.[53]

b) NSCAI Report on U.S. National Security

On November 4, 2019, the National Security Commission on Artificial Intelligence (“NSCAI”)—which was tasked by Congress to research ways to advance the development of AI for national security and defense purposes—released a highly anticipated interim report specifying five key areas in which U.S. policy can improve in order to transition AI from “a promising technological novelty into a mature technology integrated into core national security missions.”[54] The commission worked with a number of U.S. government departments and agencies including the intelligence community, academia and the private sector, as well as allied partners such as the United Kingdom, Japan, Canada and Australia. Across all five principles, NSCAI said that ethical and responsible development and deployment of AI is a top priority, and noted that it is still developing best practices for operationalizing AI technologies that are trustworthy, explainable, and free of unwanted bias. The five lines of effort are: invest in research and development; apply the technology to national security missions; train and recruit AI talent; protect and build upon U.S. technology advantages; and marshal global cooperation on artificial intelligence issues.

The commission’s preliminary conclusion is that the U.S. “is not translating broad national AI strengths and AI strategy statements into specific national security advantages.”[55] Notably, the commission reported that federal R&D funding has not kept pace with the potential of AI technologies, noting that the requested fiscal year 2020 federal funding for core AI research outside of the defense sector grew by less than 2 percent from the estimated 2019 levels.[56] Further, it noted that AI is not realizing its potential to execute core national security missions because agencies are failing to embrace the technology as a result of “bureaucratic impediments and inertia.”[57] NSCAI also criticized the shortage of AI talent in government agencies, specifically in the Department of Defense (“DoD”). It made workforce development recommendations to federal agencies, including undertaking more widespread use of AI technologies, and improving training on basic AI principles.[58] The commission asserted that the U.S. has a global technological advantage in terms of AI implementation, but also warned that China is rapidly closing the gap.[59] NSCAI recommended export controls to protect AI hardware,[60] and preservation of an open research system with U.S. academia. Finally, the commission said the U.S. should lead creation of AI norms worldwide by fostering international collaboration and establishing a network of allies dedicated to AI data sharing, R&D coordination, capacity building, and talent exchanges.

NSCAI is set to release its final report and recommendations—which will likely contain additional insights into U.S. federal policy regarding AI—in March 2021.

On October 7, 2019, the Department of Commerce Bureau of Industry and Security (“BIS”) announced that it will add 28 Chinese governmental and commercial organizations to the Entity List for engaging in or enabling activities contrary to the foreign policy interests of the United States.[61] The regulation includes China’s leading AI companies, including Sense Time, Megvii Technology, Yitu, and Dahua Technology. Companies are required to comply with the notice as of the effective date, although it includes a standard “savings clause” exempting items that are already en route as of October 9, 2019. The Secretary of Commerce stated that this action was in response to “the brutal suppression of ethnic minorities within China[.]”[62]

On January 3, 2020, BIS also announced its first unilateral control on a specific application of AI in software that automates certain data analysis of geospatial imagery data.[63] The United States has previously imposed controls on certain enabling technologies for AI in concert with other countries that participate in multilateral export control regimes. However, this is the first new AI control imposed by BIS since it began evaluating potential controls on AI as one of among several types of emerging or foundational technologies pursuant to congressionally imposed mandate in the Export Control Reform Act of 2018.

While BIS has indicated in many public fora that it will strive to ensure that any new controls it may impose on emerging and foundational technologies like AI are also adopted in peer countries that participate in the Wassenaar Arrangement, among other multilateral export regimes, the development of international consensus around specific controls often requires years of outreach and negotiation. Likely due to the significant national security-related concerns associated with development of AI-enabled, automated geospatial data analysis software, BIS opted to act unilaterally now. The specific software now subject to controls is described under Export Control Classification Number (ECCN) 0D521. Effective immediately, all exports of the software to countries worldwide (except Canada) and will now require an individual license from BIS. Moreover, releases of the software in source code to non-U.S. persons, for example, non-U.S. person employees, also require licensing. The only exception for the new license requirement is for exports of the software when transferred by or to a department or agency of the U.S. Government. In addition to these new licensing requirements, BIS’s control of the software under the new ECCN makes the software a kind of critical technology for the purposes of foreign investment review by the Committee on Foreign Investment in the United States (“CFIUS”).

III. EU POLICY & REGULATORY DEVELOPMENTS

In 2019, the European Union (“EU”) announced that it was preparing comprehensive legislation to govern AI, took steps to demonstrate its commitment toward the advancement of AI technology through funding,[64] while simultaneously pressing for companies and governments to develop ethical applications of AI. As we have addressed previously,[65] given the stringent requirements of the European General Data Protection Regulation (“GDPR”), future EC regulations are likely to stand in contrast to the current U.S. “light-touch” regulatory approach and could have a significant impact on companies developing or operating AI products within the EU. Given that the U.S. and China currently lead the global AI race in terms of technological advancement, the “regulate-first” approach of the European Union (“EU”) has led to concerns that it will impede innovation within the EU.[66]

A. EC Focus on Comprehensive AI Regulation

In mid-2019, the new president of the European Commission (“EC”), Ursula von der Leyen, unveiled her five-year policy agenda and promised to put forward legislation “for a coordinated European approach on the human and ethical implications of AI” by March 2020.[67]

In a speech at the European Parliament on November 27, 2019, von der Leyen said that she was in favor of AI-focused legislation similar to the GDPR.[68] The Commission is also likely to draw on the work of its high-level expert group on AI, which outlined a series of principles earlier this year aimed at ensuring companies deploy artificial intelligence in a way that is fair, safe and accountable. In January 2020, a leaked draft of a white paper noted that the EC was considering a five-year ban on the use of facial recognition technology in public spaces, although recent press reports indicate that the EC has since scrapped the possibility of a ban.[69] The draft also suggested that the EU’s executive body is in fact leaning towards tweaks of existing rules and sector/application-specific risk assessments and requirements, rather than blanket sectoral requirements or bans.[70] The proposal also emphasizes the need for an oversight governance regime to ensure rules are followed—though the Commission suggested leaving Member States to choose whether to rely on existing governance bodies for this task or create new ones dedicated to regulating AI. The revised proposal, part of a package of measures to address the challenges of AI, could still be amended before the Commission presents its plan on February 19, 2020.

On the basis of these statements, we anticipate that the AI legislation will:

(1) be comprehensive and sweeping in nature, aiming to address fundamental questions at a more abstract level (similar to the GDPR);

(2) be focused on individual rights (including information rights) and require GDPR-style impact assessments to ensure AI systems do not perpetuate discrimination or violate fundamental rights;

(3) address government funding of research, workplace training and the availability of public data;

(4) require, like some U.S. states—notably California—that any chatbot or virtual assistant interacting with individuals will need to disclose that it is not a human, and create enhanced requirements for transparency as to the use of data and the bases for decisions or recommendations to avoid unintended bias or disparate impact; and

(5) contain rules for accountability, mitigation of bias and discrimination, liability and transparency throughout the entire life cycle of a product or service.[71]

The legislative initiative is part of a bigger effort to secure a competitive advantage and to increase public and private investment in AI to €20 billion per year.[72] A key challenge for the new president of the EC von der Leyen will be to grow investment, data, and talent required to develop AI and accelerate its adoption, and to create an innovation-friendly regulatory environment across the EU.

B. Ethics Guidelines for Trustworthy AI

Another focus of regulatory activity within the EU and individual EU Member States has been the development of ethical considerations in the use of AI. In connection with the implementation of the GDPR in 2018, in April 2019, the EC released a report from its “High-Level Expert Group on Artificial Intelligence” (“AI HLEG”): the EU “Ethics Guidelines for Trustworthy AI” (“Ethics Guidelines”).[73] The Ethics Guidelines lay out seven ethical principles “that must be respected in the development, deployment, and use of AI systems”:

(1) Human Agency and Oversight: AI systems should enable equitable societies by supporting human agency and fundamental rights, and not decrease, limit or misguide human autonomy.

(2) Robustness and Safety: Trustworthy AI requires algorithms to be secure, reliable and robust enough to deal with errors or inconsistencies during all life cycle phases of AI systems.

(3) Privacy and Data Governance: Citizens should have full control over their own data, while data concerning them will not be used to harm or discriminate against them.

(4) Transparency: The traceability of AI systems should be ensured.

(5) Diversity, Non-Discrimination and Fairness: AI systems should consider the whole range of human abilities, skills and requirements, and ensure accessibility.

(6) Societal and Environmental Well-Being: AI systems should be used to enhance positive social change and enhance sustainability and ecological responsibility.

(7) Accountability: Mechanisms should be put in place to ensure responsibility and accountability for AI systems and their outcomes.

In addition, the Ethics Guidelines highlight the importance of implementing a “large-scale pilot with partners” and of “building international consensus for human-centric AI,”[74] and contain an “assessment list” which operationalizes the ethical principles and offers guidance to implement them in practice.[75] Along with the release of the Ethics Guidelines, the EC initiated a pilot phase of guideline implementation to assess the practical implementation of the assessment list and to gather feedback on how it can be improved.[76] Following the end of the piloting phase in December 2019, the AI HLEG will evaluate the feedback received and propose a revised version of the assessment list to the EC in early 2020.[77]

The EU also intends to “continue to play an active role in international discussions and initiatives including the G7 and G20.”[78] While the Guidelines do not appear to create any binding regulation on stakeholders in the EU, their further development and evolution will likely shape the final version of future regulation throughout the EU and therefore merits continued attention.

C. German Data Ethics Commission Report

On October 23, 2019, Germany’s Data Ethics Commission (“Ethics Commission”) released a landmark 240-page report containing 75 recommendations for regulating data, algorithmic systems and AI.[79] Consistent with EC President Ursula von der Leyen’s recent remarks, the report suggests that EU regulation of AI may mirror the approach espoused in the GDPR—broad in scope, focused on individual rights and corporate accountability, and “horizontally” applicable across industries, rather than specific sectors.[80] Expanding on the EU’s non-binding “Ethics Guidelines for Trustworthy AI,” the commission concludes that “regulation is necessary, and cannot be replaced by ethical principles.”[81]

The report creates a blueprint for the implementation of binding legal rules for AI—nominally both at national and EU level—on a sliding scale based on the risk of harm across five levels of algorithmic systems, with a focus on the degree of potential harm rather than differentiating between specific use cases. While systems posing a negligible or low likelihood of harm would not require any new regulatory obligations, those with at least “some” potential for harm would be subject to a mandatory labeling scheme that indicates where and how algorithms are being used within the system, and a risk assessment that evaluates the system’s effect on privacy rights, self-determination, bodily or personal integrity, assets and ownership rights, and discrimination, among other factors. For systems that curate content based on user data, such as personalized pricing algorithms, the commission recommends prior authorization by supervisory institutions, and heightened oversight (such as live monitoring) and transparency obligations systems with “regular or significant potential for harm,” which include determinations about consumer creditworthiness. The commission recommended a full or partial ban on systems with an “untenable potential for harm.”[82]

Of particular relevance to companies deploying AI software, the Ethics Commission recommends that measures be taken against “ethically indefensible uses of data,” such as “total surveillance, profiling that poses a threat to personal integrity, the targeted exploitation of vulnerabilities, addictive designs and dark patterns, methods of influencing political elections that are incompatible with the principle of democracy, vendor lock-in and systematic consumer detriment, and many practices that involve trading in personal data.”[83] The Ethics Commission also recommends that human operators of algorithmic systems be held vicariously liable for any harm caused by autonomous technology, and calls for an overhaul of existing product liability and strict liability laws as they pertain to algorithmic products and services.[84]

While the report’s pro-regulation approach is a counterweight to the “light-touch” regulation favored by the U.S. government, the Ethics Commission takes the view that, far from impeding private sector innovation, regulation can provide much-needed certainty to companies developing, testing, and deploying innovative AI products.[85] Certainly, the Ethics Commission’s guiding principles—among them the need to ensure “the human-centred and value-oriented design of technology”[86]—reinforce that European lawmakers are likely to regulate AI development comprehensively and decisively. While it remains to be seen to what extent the forthcoming draft EU legislation will adopt the commission’s recommendations, all signs point to a sweeping regulatory regime that could significantly impact technology companies active in the EU.

IV. REGULATION OF AI TECHNOLOGIES AND ALGORITHMS

As the use of AI expands into different sectors and the need for data multiplies, legislation that traditionally has not focused on AI is starting to have a growing impact on AI technology development. Defining and achieving an ethical approach to AI decision-making has been at the forefront of policy discussions relating to the private sector for some time, and the deep learning community has responded with a wave of investments and initiatives focusing on processes designed to assess and mitigate bias and disenfranchisement[87] at risk of becoming “baked in and scaled” by AI systems.[88] Such discussions are now becoming more urgent and nuanced with the increased availability of AI decision-making tools allowing government decisions to be delegated to algorithms to improve accuracy and drive objectivity, directly impacting democracy and governance.[89] Over the past year, we have seen those discussions evolve into tangible and impactful legislative proposals and concrete regulations in the data regulation space and, notably, several outright technology bans.[90]

A. Algorithmic Accountability

In 2019, a number of federal bills addressing algorithmic accountability and transparency hinted at a shift in Washington’s stance amid growing public awareness of AI’s potential to create bias or harm certain groups.[91] While the proposed legislation remains in its early stages, it is indicative of the government’s increasingly bold engagement with technological innovation and the regulation of AI, and companies operating in this space should remain alert to both opportunities and risks arising out of federal legislative and policy developments—particularly the increasing availability of public-private partnerships—during 2020.

1. H.R. 153

In February 2019, the House introduced Resolution 153 , with the intent of “[s]upporting the development of guidelines for ethical development of artificial intelligence” and emphasizing the “far-reaching societal impacts of AI” as well as the need for AI’s “safe, responsible, and democratic development.”[92] Similar to California’s adoption last year of the Asilomar Principles[93] and the OECD’s recent adoption of five “democratic” AI principles,[94] the House Resolution provides that the guidelines must be consonant with certain specified goals, including “transparency and explainability,” “information privacy and the protection of one’s personal data,” “accountability and oversight for all automated decisionmaking,” and “access and fairness.” This Resolution puts ethics at the forefront of policy, which differs from other legislation that considers ethics only as an ancillary topic. Yet, while this resolution signals a call to action by the government to come up with ethical guidelines for the use of AI technology, the details and scope of such ethical regulations remain unclear.

2. Algorithmic Accountability Act

On April 10, 2019, a number of Senate Democrats introduced the “Algorithmic Accountability Act,” which “requires companies to study and fix flawed computer algorithms that result in inaccurate, unfair, biased or discriminatory decisions impacting Americans.”[95] Rep. Yvette D. Clarke (D-NY) introduced a companion bill in the House.[96] The bill stands to be the United States Congress’s first serious foray into the regulation of AI and the first legislative attempt in the United States to impose regulation on AI systems in general, as opposed to regulating a specific technology area. The bill reflects a step back from the previously favored approach of industry self-regulation, since it would force companies to actively monitor use of any potentially discriminatory algorithms. Although it does not provide for a private right of action or enforcement by state attorneys general, it would give the Federal Trade Commission the authority to enforce and regulate these audit procedures and requirements. Further congressional action on this subject can certainly be anticipated.

The bill casts a wide net, such that many technology companies would find common practices to fall within the purview of the Act. The Act would not only regulate AI systems but also any “automated decision system,” which is broadly defined as any “computational process, including one derived from machine learning, statistics, or other data processing or artificial intelligence techniques, that makes a decision or facilitates human decision making, that impacts consumers.”[97] For processes within the definition, companies would be required to audit for bias and discrimination and take corrective action to resolve these issues, when identified. The bill would allow regulators to take a closer look at any “[h]igh-risk automated decision system”—those that involve “privacy or security of personal information of consumers[,]” “sensitives aspects of [consumers’] lives, such as their work performance, economic situation, health, personal preferences, interests, behavior, location, or movements[,]” “a significant number of consumers regarding race [and several other sensitive topics],” or “systematically monitors a large, publicly accessible physical place[.]”[98] For these “high-risk” topics, regulators would be permitted to conduct an “impact assessment” and examine a host of proprietary aspects relating to the system. Additional regulations will be needed to give these key terms meaning but, for now, the bill is a harbinger for AI regulation that identifies key areas of concern for lawmakers.

Although the bill still faces an uncertain future, if it is enacted, businesses would face a number of challenges, not least significant uncertainty in defining and, ultimately, seeking to comply with the proposed requirements for implementing “high risk” AI systems and utilizing consumer data, as well as the challenges of sufficiently explaining to the FTC the operation of their AI systems. Moreover, the bill expressly states that it does not preempt state law—and states that have already been developing their own consumer privacy protection laws would likely object to any attempts at federal preemption—potentially creating a complex patchwork of federal and state rules.[99] At a minimum, companies operating in this space should certainly anticipate further congressional action on this subject in the near future, and proactively consider how their own “high-risk” systems may raise concerns related to bias.

3. Bot Disclosure and Accountability Act (S. 2125)

The Bot Disclosure and Accountability Act, first introduced on June 25, 2018 and reintroduced on July 16, 2019, mandates that the FTC come up with regulations that force digital platforms to publicly disclose their use of an ‘automated software program or process intended to replicate human activity online’.[100] It also prohibits political candidates or parties from using these automated software programs in order to share or disseminate any information targeting political elections. The Act hands the task of defining ‘automated software program’ to the FTC, which leaves wide latitude in interpretation beyond the narrow bot purpose for which the bill is intended.

At the state level, California passed a bill in September 2018, the ‘Bolstering Online Transparency Act’,[101] which was the first of its kind and (similar to the federal bot bill) is intended to combat malicious bots operating on digital platforms. This state law does not attempt to ban bots outright, but requires companies to disclose whether they are using a bot to communicate with the public on their internet platforms. The law went into effect on July 1, 2019.

4. Filter Bubble Transparency Act (S. 2763)

On October 31, 2019 a bipartisan group of senators introduced the Filter Bubble Transparency Act, the first substantive federal bill aimed at regulating algorithmic control of content on internet platforms. If enacted, the bill would require large-scale internet platforms to provide greater transparency to consumers by providing clear notice on the use, and enabling consumers to opt out, of personalized content curated by “opaque” algorithms so that they can “engage with a platform without being manipulated by algorithms driven by user-specific data”[102] and “simply opt out of the filter bubble.”[103] “Filter bubble” refers to a zone of potential manipulation that exists within algorithms that curate or rank content in internet platforms based on user-specific data, potentially creating digital “echo chambers.”[104]

The proposed legislation covers “any public-facing website, internet application, or mobile application,” such as social network sites, video sharing services, search engines and content aggregation services,[105] and generally would prohibit the use of opaque algorithms on platforms without those platforms having first provided notice in a “clear, conspicuous manner on the platform whenever the user interacts with an opaque algorithm for the first time.” The term “opaque algorithm” is defined as “an algorithmic ranking system[106] that determines the order or manner that information is furnished to a user on a covered internet platform based, in whole or part, on user-specific data that was not expressly provided by the user to the platform” in order to interact with it.[107] Examples of “user-specific” data include the user’s history of web searches and browsing, geographical locations, physical activity, device interaction, and financial transactions.[108] Conversely, data that was expressly provided to the platform by the user for the purpose of interacting with the platform—such as search terms, saved preferences, an explicitly entered geographical location or the user’s social media profiles[109]—is considered “user-supplied.”

Additionally, the bill requires that users be given the option to choose to view content based on “input-transparent algorithms,” a purportedly generic algorithmic ranking system that “does not use the user-specific data of a user to determine the order or manner that information is furnished to such user on a covered platform,”[110] and be able to easily switch between the opaque and the input-transparent versions.[111] By way of example, Sen. Marsha Blackburn (R-TN), another co-sponsor of the bill, explained that “this legislation would give consumers the choice to decide whether they want to use the algorithm or view content in the order it was posted.”[112] However, there is nothing in the bill that would require platforms to disclose the use of algorithms unless they are using hyper-personal “user-specific” data for customization, and even “input-transparent” algorithms using “user-supplied” data would not necessarily show content in chronological order. Nor would platforms be required to disclose any source code or explain how the algorithms used work. As drafted, the bill’s goals of providing transparency and protecting consumers from algorithmic manipulation by “opting out” of personalized content appear to be overstated, and lawmakers will need to grapple with the proposed definitions to clarify the scope of the bill’s provisions.[113]

Like the Algorithmic Accountability Act, the bill is squarely targeted at “Big Tech” platforms—it would not apply to platforms wholly owned, controlled and operated by a person that did not employ more than 500 employees in the past six months, averaged less than $50 million in annual gross receipts, and annually collects or processes personal data of less than a million individuals.[114] Violations of the Act would be enforced with civil penalties by the Federal Trade Commission (“FTC”) but, unlike the Algorithmic Accountability Act, the bill does not grant state attorneys general the right to bring civil suits for violations, nor expressly state that its provisions do not preempt state laws.[115]

B. Facial Recognition Software

Biometric surveillance, or “facial recognition technology,” has emerged as a lightning rod for public debate regarding the risk of improper algorithmic bias and data privacy concerns. Until recently, there were few if any laws or guidelines governing the use of facial recognition technology. Amid widespread fears that the current state of the technology is not sufficiently accurate or reliable to avoid discrimination, regulators have seized the opportunity to act in the AI space—proposing and passing outright bans on the use of facial recognition technology with no margin for discretion or use case testing while a broader regulatory approach develops and the technology evolves.[116] This tentative consensus stands in stark contrast to the generally permissive approach to the development of AI systems in the private sector to date. While much of the regulatory activity to date has been at the local level, momentum is also building for additional regulatory actions at both the state and federal levels.

1. Federal Regulation

The federal government has indicated a willingness to consider a nationwide ban on facial recognition technology, or at least to enact stringent regulations. A bill introduced in Congress in March 2019 (S. 847, “Commercial Facial Recognition Privacy Act of 2019”) would ban users of commercial face recognition technology from collecting and sharing data for identifying or tracking consumers without their consent, although it does not address the government’s uses of the technology.[117] With few exceptions, the bill would require facial recognition technology available online to be made accessible for independent third-party testing “for accuracy and bias.” The bill remains pending and has been referred to the Committee on Commerce, Science, and Transportation.

Several other federal bills on facial recognition technology have been proposed. H.R. 3875 was introduced on July 22, 2019, to “prohibit Federal funding from being used for the purchase or use of facial recognition technology.”[118] On July 25, 2019, Representatives Yvette Clark (D-NY), Ayanna Pressley (D-Mass.) and Rashida Tlaib (D-Mich.) introduced the “No Biometric Barriers to Housing Act.” If passed, the bill would prohibit facial recognition in public housing units that receive Department of Housing and Urban Development (“HUD”) funding. It would also require HUD to submit a report on facial recognition and its impacts on public housing units and tenants.[119] Also on July 25, 2019, the Facial, Analysis, Comparison, and Evaluation (“FACE”) Protection Act of 2019 (H.R. 4021) was introduced to prohibit a federal agency from applying “facial recognition technology to any photo identification issued by a State or the Federal Government or any other photograph otherwise in the possession of a State or the Federal Government unless the agency has obtained a Federal court order determining that there is probable cause for the application of such technology.”[120]

The House Committee on Oversight and Reform has held several hearings on transparency regarding government use cases, at which Committee members voiced strong bipartisan support for providing transparency and accountability to the use of facial recognition technology.[121] To date, the group is continuing to work on legislation that could regulate the use of facial recognition by the private sector, federal government, and law enforcement. On January 15, 2020, the House Oversight and Reform Committee held its third hearing in less than a year about facial recognition, this time to explore its use in the private sector.[122]

2. State and Local Regulations

In 2019, lawmakers in numerous states introduced bills to ban or delay the use of facial recognition technology by government agencies or the private sector. In September 2019, California lawmakers passed legislation (A.B. 1215) which places a three-year moratorium on any facial recognition technology used in police body cameras beginning January 1, 2020.[123] The bill by Assemblyman Phil Ting (D-San Francisco), which was co-sponsored by the ACLU, was signed into law by Governor Newsom on October 8, 2019.[124] Previously, the ACLU had run demonstrations using facial-recognition technology which falsely flagged 26 California lawmakers as matching arrest photos.[125]

The language of A.B. 1215 states that using biometric surveillance violates constitutional rights because it is the “functional equivalent” of requiring people to carry identification at all times.[126] The new law further regulates the collection of personal information, sounds in California’s concern for overly broad collection of information, and may influence modifications to the California Consumer Privacy Act 2018 (“CCPA”) regarding facial recognition (such as A.B. 1281, which would require businesses to give conspicuous notices where facial recognition technology is employed).

Amid increasing public concern about the technology operating in public spaces, 2019 also saw a string of efforts by various cities in the U.S. to ban the use of facial recognition technology by law enforcement.[127] Oakland City Council passed an ordinance to ban its use by city police and other government departments, joining San Francisco, California and Somerville, Massachusetts who had already enacted similar bans.[128] Berkeley City Council also adopted a ban at a meeting in mid-October 2019.[129]

C. Deepfake technology

A new AI application called “deepfakes” is raising a set of challenging policy, technology, and legal issues. Deepfake technology is used to combine and superimpose existing images and videos onto source images or videos—creating new “synthetic” images or videos—by using a machine learning technique known as a generative adversarial network (“GAN”), a deep neural net architecture comprised of two nets, pitting one against the other (the “adversarial”). Since GANs can learn to mimic any distribution of data (images, music, speech, or text), the applications of deepfake technology are vast. Prompted by increased public concern over the potential impact of the technology on everything from cybersecurity to electoral manipulation, tentative federal bills intended to regulate deepfakes have emerged over the past several months, while state legislatures have already reacted by banning certain deepfake applications.[130]

1. Federal Regulatory Efforts

In September 2018, Reps. Adam Schiff (D-Calif.), Stephanie Murphy (D-Fla.) and Carlos Curbelo (R-Fla.) sent a letter to the Director of National Intelligence to warn of potential risks relating to deepfakes.[131] The lawmakers cautioned that “[d]eep fakes have the potential to disrupt every facet of our society and trigger dangerous international and domestic consequences . . . [a]s with any threat, our Intelligence Community must be prepared to combat deep fakes, be vigilant against them, and stand ready to protect our nation and the American people.”[132] In the wake of a June 2019 hearing by the House Permanent Select Committee on Intelligence on the national security challenges of artificial intelligence, manipulated media, and deepfake technology, both the House and the Senate introduced legislation to regulate GANs. At present, however, the bills appear to do very little to restrict the use of deepfake technology, suggesting that Congress remains in “learning mode.”

On July 9, 2019, Sen. Rob Portman (R-OH) introduced the “Deepfake Report Act” (S. 2065), which would require the Department of Homeland Security to submit five annual reports to Congress on the state of the “digital content forgery” technology and evaluate available methods of detecting and mitigating threats.[133] The reports will include assessments of how the technology can be used to harm national security as well as potential counter measures. The bill defines digital content forgery as “the use of emerging technologies, including artificial intelligence and machine learning techniques, to fabricate or manipulate audio, visual, or text content with the intent to mislead.” The bipartisan bill was passed in the Senate by unanimous consent on October 25 and is currently before the House Committee on Energy and Commerce, which is reviewing the same-named companion bill, H.R. 3600.[134]

In the House, H.R. 3230 (“Defending Each and Every Person from False Appearances by Keeping Exploitation Subject to Accountability Act” or the “DEEPFAKES Act”) was introduced by Rep. Clarke (D-NY-9) on June 12, 2019.[135] It would require any “advanced technological false personation record” to be digitally watermarked. The watermark would be required to “clearly identifying such record as containing altered audio or visual elements.” The bill has been referred to the Subcommittee on Crime, Terrorism, and Homeland Security.

On September 17, 2019, Rep. Anthony Gonzalez (R-OH) introduced the “Identifying Outputs of Generative Adversarial Networks Act” (H.R. 4355), which would direct both the National Science Foundation and NIST to support research on deepfakes to accelerate the development of technologies that could help improve their detection, to issue a joint report on research opportunities with the private sector, and to consider the feasibility of ongoing public and private sector engagement to develop voluntary standards for the outputs of GANs or comparable technologies.[136]

2. State Regulatory Efforts

In the wake of a June 2019 hearing by the House Permanent Select Committee on Intelligence on the national security challenges of artificial intelligence, manipulated media, and deepfake technology, both the House and the Senate introduced legislation to regulate deepfakes.

While those bills remains pending, California has taken action to restrict the specific use of deepfakes to influence elections and non-consensual pornographic deepfakes. On October 3, 2019 California’s Gov. Newsom signed a bill (A.B. 730) banning anyone “from distributing with actual malice materially deceptive audio or visual media of the candidate” within 60 days of an election with the intent to injure the candidate’s reputation or deceive a voter into voting for or against the candidate.[137] This measure exempts print and online media and websites if that entity clearly discloses that the deepfake video or audio file is inaccurate or of questionable authenticity. On October 3, Gov. Newsom also signed a bill (A.B. 602) banning pornographic deepfakes made without consent of the person depicted, creating a private right of action.[138] The law excepts “[c]ommentary, criticism, or disclosure that is otherwise protected by the California Constitution or the United States Constitution.”

These laws may signal state regulators’ willingness to quickly regulate other controversial AI applications going forward. It will remain to be seen whether these laws will be challenged and whether they will pass constitutional muster. Regardless, the use and proliferation of deepfakes will likely face greater legal and regulatory scrutiny at both federal and state level going forward, and may impact technology platforms which permit users to upload, share or link content.

D. Autonomous Vehicles

1. Federal Developments

There was a flurry of legislative activity in Congress in 2017 and early 2018 towards a national regulatory framework. The U.S. House of Representatives passed the Safely Ensuring Lives Future Deployment and Research In Vehicle Evolution (SELF DRIVE) Act (H.R. 3388)[139] by voice vote in September 2017, but its companion bill (the American Vision for Safer Transportation through Advancement of Revolutionary Technologies (AV START) Act (S. 1885)),[140] stalled in the Senate as a result of holds from Democratic senators who expressed concerns that the proposed legislation remains immature and underdeveloped in that it “indefinitely” preempts state and local safety regulations even in the absence of federal standards.[141] Federal regulation of autonomous vehicles (“AVs”) has so far faltered in the new Congress, as SELF DRIVE Act and the AV START Act have not been re-introduced since expiring with the close of the 115th Congress.[142]

In 2019, federal lawmakers have demonstrated renewed interest in a comprehensive AV bill aimed at speeding up the adoption of autonomous vehicles and deploying a regulatory framework. In July 2019, the House Energy and Commerce Committee and Senate Commerce Committee sought stakeholder input from the self-driving car industry in order to draft a bipartisan and bicameral AV bill, prompting stakeholders to provide feedback to the committees on a variety of issues involving autonomous vehicles, including cybersecurity, privacy, disability access, and testing expansion.[143] Moreover, several federal agencies have announced proposed rulemaking to facilitate the integration of autonomous vehicles onto public roads. And while federal regulations are lagging behind, legislative activity at the state and local level is stepping up to advance integration of autonomous vehicles in the national transportation system and local infrastructure.

In the meantime, AVs continue to operate under a complex patchwork of state and local rules, with federal oversight limited to the U.S. Department of Transportation’s (“DoT”) informal guidance. In January 2020, the DoT published updated guidance for the regulation of the autonomous vehicle industry, “Ensuring American Leadership in Automated Vehicle Technologies” or “AV 4.0.”[144] The guidance builds on the AV 3.0 guidance released in October 2018, which introduced guiding principles for AV innovation for all surface transportation modes, and described the DoT’s strategy to address existing barriers to potential safety benefits and progress.[145] AV 4.0 includes 10 principles to protect consumers, promote markets and ensure a standardized federal approach to AVs. In line with previous guidance, the report promises to address legitimate public concerns about safety, security, and privacy without hampering innovation, relying strongly on the industry self-regulating. However, the report also reiterates traditional disclosure and compliance standards that companies leveraging emerging technology should continue to follow.

During 2019, several federal agencies announced proposed rule-making to facilitate the integration of autonomous vehicles onto public roads. In May 2019, in the wake of a petition filed by General Motors requesting temporary exemption from Federal Motor Vehicle Safety Standards (FMVSSs) which require manual controls or have requirements that are specific to a human driver,[146] NHTSA announced that it was seeking comments about the possibility of removing ‘regulatory barriers’ relating to the introduction of automated vehicles in the United States.[147] It is likely that regulatory changes to testing procedures (including preprogrammed execution, simulation, use of external controls, use of a surrogate vehicle with human controls and technical documentation) and modifications to current FMVSSs (such as crashworthiness, crash avoidance and indicator standards) will be finalized in 2021.

2. State Developments

State regulatory activity has continued to accelerate, adding to the already complex patchwork of regulations that apply to companies manufacturing and testing autonomous vehicles. State regulations vary significantly, ranging from allowing testing under certain specific and confined conditions to the more extreme, which allow for testing and operating AVs with no human passenger behind the wheel. Recognizing that AVs and vehicles with semi-autonomous components are already being tested and deployed on roads amid legislative gridlock at the federal level, 44 states and the District of Columbia have enacted autonomous vehicle legislation. In 2019 alone, 25 new bills were enacted in 25 states, and a further 56 remain pending.[148] Increasingly, there are concerns that states may be racing to cement their positions as leaders in AV testing in the absence of a federal regulatory framework by introducing increasingly permissive bills that allow testing without human safety drivers.[149]

Some states are explicitly tying bills to federal guidelines in anticipation of congressional action. On April 2, 2019, D.C. lawmakers proposed the Autonomous Vehicles Testing Program Amendment Act of 2019, which would set up a review and permitting process for autonomous vehicle testing within the District Department of Transportation. Companies seeking to test self-driving cars in the city would have to provide an array of information to officials, including— for each vehicle it plans to test—safety operators in the test vehicles, testing locations, insurance, and safety strategies.[150] Crucially, it would require testing companies to certify that their vehicles comply with federal safety policies; share with officials data on trips and any crash or cybersecurity incidents; and train operators on safety.[151]

On April 12, 2019, the California DMV published proposed autonomous vehicle regulations that allow the testing and deployment of autonomous motor trucks (delivery vehicles) weighing less than 10,001 pounds on California’s public roads.[152] The DMV held a public hearing on May 30, 2019, at its headquarters in Sacramento to gather input and discuss the regulations. The DMV’s regulations continue to exclude the autonomous testing or deployment of vehicles weighing more than 10,001 pounds. In the California legislature, two new bills related to autonomous vehicles were introduced: S.B. 59[153] would establish a working group on autonomous passenger vehicle policy development while S.B. 336[154] would require transit operators to ensure certain automated transit vehicles are staffed by employees.

On June 13, 2019, Florida Governor Ron DeSantis signed into law C.S./H.B. 311, which establishes a statewide statutory framework, permits fully automated vehicles to operate on public roads, and removes obstacles that hinder the development of self-driving cars.[155] In Oklahoma, Governor Kevin Stitt signed legislation (S.B. 365) restricting city and county governments from legislating autonomous vehicles, ensuring that such legislation would be entirely in the hands of state and federal lawmakers.[156] Pennsylvania, which last year passed legislation creating a commission on “highly automated vehicles,” has proposed a bill that would authorize the use of an autonomous shuttle vehicle on a route approved by the Pennsylvania Department of Transportation (H.B. 1078).[157]

Given the fast pace of developments and tangle of applicable rules, it is essential that companies operating in this space stay abreast of legal developments in states as well as cities in which they are developing or testing autonomous vehicles, while understanding that any new federal regulations may ultimately preempt states’ authorities to determine, for example, safety policies or how they handle their passengers’ data.

E. Data Privacy

While not strictly focused on artificial intelligence technologies, a number of state and federal developments in the area of data privacy are noteworthy, given the central importance of access to large quantities of data (often including personal and private data) to the successful development and operation of many AI systems.[158]

1. Voter Privacy Act of 2019

In July 2019, California Senator Dianne Feinstein introduced the Voter Privacy Act of 2019, which is currently before the Senate Committee on Rules and Administration.[159] As introduced, the Act will give voters certain rights with regard to their personal data collected in connection with voter information. In particular, the Act provides notice rights, rights of access, deletion rights, and rights to prohibit transfer or targeting through use of the data. The stated purpose of the Act is to put an end to the manipulation and misdirection of voters through the use of their personal data, and the Act would be monitored by the Federal Election Commission. Obviously, for companies collecting voter information as part of the data processed by AI systems, the Act could add a number of significant compliance requirements should it ultimately pass.

2. California Consumer Privacy Act (“CCPA”)

A series of amendments to the California Consumer Privacy Act (“CCPA”) were signed into effect by the Governor in early October.[160] Some of these amendments may prove significant to certain businesses; such as A.B. 25, which provides a one-year carve-out of the personal information of employees from personal information that would otherwise fall under the requirements of the CCPA. Similarly, A.B. 1355 creates a one-year carve-out of certain personal information that is collected as part of purely business-to-business communications, which may also help alleviate concerns about how to handle personal information necessarily acquired in a business context. In addition to the amendments, the California Attorney General’s Office released a series of proposed regulations for implementing the requirements of the CCPA, and initiated a period in which they will solicit public comments before making any final changes putting the regulations into force and effect.[161] The proposed regulations generally set out guidance for how businesses should implement the notice provisions of the CCPA, procedural steps for implementing consumer rights provisions and data collection requirements, as well as provide some clarification of the CCPA’s non-discrimination provisions. The CCPA has been described as one of the most stringent state privacy laws and will affect AI technologies that are driven by personal data and companies who utilize or develop such technologies.

3. California “Anti-Eavesdropping Act”

On May 29, 2019 the California State Assembly passed a bill (A.B. 1395) requiring manufacturers of ambient listening devices like smart speakers to receive consent from users before retaining voice recordings, and banning manufacturers from sharing command recordings with third parties. The bill is currently pending in the State Senate.[162]

F. Intellectual Property

Intellectual property issues related to AI have also been at the forefront of the new technology, as record numbers of U.S. patent applications involve a form of machine learning component. In January 2019, the United States Patent and Trademark Office (“USPTO”) released revised guidance relating to subject matter eligibly for patents and on the application of 35 U.S.C. § 112 on computer implemented inventions. On the heels of that guidance, on August 27, 2019, the USPTO published a request for public comment on several patent-related issues regarding AI inventions.[163] The request for comment posed 12 questions covering several topics from “patent examination policy to whether new forms of intellectual property protection are needed.” The questions included topics such whether patent laws, which contemplate only human inventors, should be amended to allow entities other than a human being to be considered an inventor.[164] The commenting period was extended until November 8, 2019, and many of the comments submitted argue that ownership of patent rights should remain reserved for only natural or juridical persons.[165]

On December 13, 2019, the World Intellectual Property Organization (“WIPO”) published a draft issue paper on IP policy and AI, and requested comments on several areas of IP, including patents and data, and, similarly to the USPTO before it, with regard to issues of inventorship and ownership.[166] The commenting period is set to end on February 14, 2020.

G. Law Enforcement

Increasingly, algorithms are also being used at every stage of criminal proceedings, from gathering evidence to making sentencing and parole recommendations. H.R. 4368, the “Justice in Forensic Algorithms Act of 2019,” was introduced in the House on September 17, 2019, would prohibit the use of trade secrets privileges to prevent defense access to the source code of proprietary algorithms used as evidence in criminal proceedings, and require that the Director of NIST establish a program to provide for the creation and maintenance of standards for the development and use of computational forensic software (“Computational Forensic Algorithm Standards”) to protect due process rights.[167] The standards would address underlying scientific principles and methods, an assessment of disparate impact on the basis of demographic features such as race or gender, requirements for testing and validating the software and for publicly available documentation, and requirements for reports that are provided to defendants by the prosecution documenting the use and results of computational forensic software in individual cases (e.g., source code).[168]

Police departments often use predictive algorithms for various functions, such as to help identify suspects. While such technologies can be useful, there is increasing awareness building with regard to the risk of biases and inaccuracies.[169] Private groups, localities, states, and Congress have reacted to concerns fomented by AI applied to policing. In a paper released on February 13, 2019, researchers at the AI Now Institute, a research center that studies the social impact of artificial intelligence, found that police across the United States may be training crime-predicting AIs on falsified “dirty” data,[170] calling into question the validity of predictive policing systems and other criminal risk-assessment tools that use training sets consisting of historical data.[171] In some cases, police departments had a culture of purposely manipulating or falsifying data under intense political pressure to bring down official crime rates. In New York, for example, in order to artificially deflate crime statistics, precinct commanders regularly asked victims at crime scenes not to file complaints. In predictive policing systems that rely on machine learning to forecast crime, those corrupted data points become legitimate predictors, creating “a type of tech-washing where people who use these systems assume that they are somehow more neutral or objective, but in actual fact they have ingrained a form of unconstitutionality or illegality.”[172]

A Utah law (H.B. 57) requiring that law enforcement obtain a warrant before accessing any person’s electronic data went into effect in May 2019.[173] The law reflects a legislative recognition of individual privacy rights, and we will continue to closely watch this space and the extent to which its approach may be replicated in other state legislatures. Other efforts have been more limited in scope and focused only on certain AI applications, like facial recognition.[174] Beyond policy and advocacy, some groups have turned to the courts. The majority of these efforts sound in FOIA attempts to understand how police may be using predictive systems to aid their work.[175]

H. Health Care

Unsurprisingly, the use of AI in healthcare draws some of the most exciting prospects and deepest trepidation, given potential risks.[176] As of yet, there are few regulations directed at AI in healthcare specifically, but regulators have recently acknowledged that existing frameworks for medical device approval are not well-suited to AI-related technologies. The US Food and Drug Administration (“FDA”) has proposed a specific review framework for AI-related medical devices, intended to encourage a pathway for innovative and life-changing AI technologies, while maintaining the FDA’s patient safety standards.

In April 2019, the FDA recently published a discussion paper – ’Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD)’—offering that new framework for regulating health products using AI/machine learning (“AI/ML”) software as a medical device (“SaMD”), and seeking comment.[177] The paper introduces that one of the primary benefits of using AI in an SaMD product is the ability of the product to continuously update in light of an infinite feed of real-world data. But the current review system for medical devices requires a pre-market review, and pre-market review of any modifications, depending on the significance of the modification.[178] If AI-based SaMDs are intended to constantly adjust, the FDA posits that many of these modifications will require pre-market review – a potentially unsustainable framework in its current form. The paper instead proposes an initial pre-market review for AI-related SaMDs that anticipates the expected changes, describes the methodology, and requires manufacturers to provide certain transparency and monitoring, as well as updates to the FDA about the changes that in fact resulted in accordance with the information provided in the initial review. Additional discussion and guidance is expected following the FDA’s review of the comments.

I. Financial Services

As the adoption of AI technology in the U.S. continues across a wide range of industries and the public sector, legislators are increasingly making efforts to regulate applicable data standards at federal level. On May 9, 2019, Representative Maxine Waters (D-CA) announced that the House Committee on Financial Services would launch two task forces focused on financial technology (“fintech”) and AI:[179] a task force on financial intelligence that will focus on the topics of regulating the fintech sector, and an AI task force that will focus on machine learning in financial services and regulation, emerging risks in algorithms and big data, combatting fraud and digital identification technologies, and the impact of automation on jobs in financial services.[180]

On September 24, 2019, H.R. 4476, the Financial Transparency Act of 2019, was reintroduced into Congress.[181] The bipartisan bill, which calls for the Treasury secretary to create uniform, machine-readable data standards for information reported to financial regulatory agencies,[182] has been referred to the Subcommittee on Commodity Exchanges, Energy, and Credit. By seeking to make information that is reported to financial regulatory agencies electronically searchable, the bill’s supporters aim to “further enable the development of RegTech and Artificial Intelligence applications,” “put the United States on a path towards building a comprehensive Standard Business Reporting program,” and “harmonize and reduce the private sector’s regulatory compliance burden, while enhancing transparency and accountability.”[183]

J. Labor and Hiring

Amid the acceleration in the spread of AI and automated decision-making in the public and private sector, many U.S. and multinational companies have begun to use AI to streamline and introduce objectivity into their employment process.[184] While AI presents an opportunity to eliminate bias from the hiring process, it has also been seen to introduce bias because of inadequate data underlying and powering its algorithms. Legislators are taking action to recognize the potentially vast implications of AI technology on employment and employees’ rights. As a result, 2019 saw tentative legislation at federal and state level take on an increased focus upon AI in employment and hiring.

1. AI JOBS Act of 2019

On 28 January 2019, the proposed AI JOBS Act of 2019 was introduced and, if enacted, would authorize the Department of Labor to work with businesses and education institutions in creating a report that analyses the future of AI and its impact on the American labor landscape.[185] Similar to H.R. 153, this bill indicates federal recognition of the threat the introduction of AI technology poses; however, there is no indication as to what actions the federal government might take in order to offer labor protection, and the bill has not progressed to date.

2. Workers’ Right to Training Act (S. 2468)

On September 11, 2019, Sen. Brown (D-OH) introduced S. 2468, the “Workers’ Right to Training Act,” which would require employers to provide notice and training to employees whose jobs are in danger of being changed or replaced due to technology, and for other purposes.[186] “Technology” is defined in the bill as including “automation, artificial intelligence, robotics, personal computing, information technology, and e-commerce.”[187]

3. Illinois AI Video Interview Act

Employers have begun using AI-powered interview platforms—equipped with abilities such as sentiment analysis, facial recognition, video analytics, neural language processing, machine learning and speech recognition—that are capable of screening candidates against various parameters to assess competencies, experience and personality on the basis of hundreds of thousands of data points, and rank them against other candidates based on an “employability” score.[188] However, the lack of transparency resulting from the use of proprietary algorithms to hire and reject candidates has led to some regulatory pushback.

In May 2019, the Illinois legislature unanimously passed H.B. 2557 (the “Artificial Intelligence Video Interview Act”), which governs the use of AI by employers when hiring candidates.[189] State Rep. Jaime Andrade Jr. (D), who co-sponsored the bill, noted that spoken accents or cultural differences could end up improperly warping the results of a video interview, and that people who declined to sit for the assessment could be unfairly punished by not being considered for the job.[190] On August 9, 2019, Governor J.B. Pritzker signed the Act into law, effective January 1, 2020. Under the Act, an employer using videotaped interviews when filling a position in Illinois may use AI to analyze the interview footage only if the employer:

Gives notice to the applicant that the videotaped interview may be analyzed using AI for purposes of evaluating the applicant’s fitness for the position. (A Senate floor amendment removed from the bill a requirement for written notice.)
Provides the applicant with an explanation of how the AI works and what characteristics it uses to evaluate applicants.
Obtains consent from the applicant to use AI for an analysis of the video interview.
Keeps video recordings confidential by sharing the videos only with persons whose expertise or technology is needed to evaluate the applicant, and destroying both the video and all copies within 30 days after an applicant requests such destruction.

Illinois employers using such software will need to carefully consider how they are addressing the risk of AI-driven bias in their current operations, and whether hiring practices fall under the scope of the new law, which does not define “artificial intelligence,” what level of “explanation” is required, or whether it applies to employers seeking to fill a position in Illinois regardless of where the interview takes place. While the Illinois Act currently remains the only such law to date in the U.S., companies using automated technology in recruitment should expect that the increasing use of AI technology in recruitment is likely to lead to further regulatory proposals in due course.[191]

_________________________

[1] Most AI is specific to particular domains and problems. A “general” AI can be thought of as one that can be applied to a wide variety of cross-domain activities and perform at the level of, or better than a human agent, or has the capacity to self-improve its general cognitive abilities similar to or beyond human capabilities.

[2] The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems, Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems, First Edition, IEEE, 2019, at 2, available at https://standards.ieee.org/content/ieee-standards/en/industry-connections/ec/autonomous-systems.html.

[3] Michael Guihot, Will we ever agree to just one set of rules on the ethical development of artificial intelligence?, World Economic Forum (June 17, 2019), available at https://www.weforum.org/agenda/2019/06/will-we-ever-agree-to-just-one-set-of-rules-on-the-ethical-development-of-artificial-intelligence.

[4] OECD Legal Instruments, Recommendation of the Council on Artificial Intelligence (May 21, 2019), available at https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0449.

[5] See http://www.oecd.org/going-digital/ai/oecd-aigo-membership-list.pdf for the full list.

[6] The recommendations also instruct the OECD’s Committee on Digital Economy Policy (“CDEP”) to monitor the implementation of the recommendations and report to the Council on its implementation, to develop practical guidance for implementation and to promote the OECD AI Policy Observatory, an interactive forum for exchanging information on AI policy and activities due to launch on February 27, 2020 that will include a live database of AI strategies and initiatives as well as certain AI metrics, measurements, policies and good practices. See OECD, OECD AI Policy Observatory: A platform for AI information, evidence, and policy options (Sept. 2019), available at https://www.oecd.org/going-digital/ai/about-the-oecd-ai-policy-observatory.pdf.

[7] Press release, Canada and France work with international community to support the responsible use of artificial intelligence (May 16, 2019), available at https://www.gouvernement.fr/sites/default/files/locale/piece-jointe/2019/05/23_cedrico_press_release_ia_canada.pdf.

[8] Tim Simonite, The World Has a Plan to Rein in AI—but the US Doesn’t Like It, Wired (January 6, 2020), available at https://www.wired.com/story/world-plan-rein-ai-us-doesnt-like/

[9] Supra note 2, The IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems, Ethically Aligned Design: A Vision for Prioritizing Human Well-being with Autonomous and Intelligent Systems, at 2.

[10] Id., at 17.

[11] Id., at 4.

[12] Autonomous Weapons that Kill Must be Banned, Insists UN Chief, UN News (Mar. 25, 2019), available at https://news.un.org/en/story/2019/03/1035381.

[13] Japan Pledges No AI “Killer Robots,” MeriTalk (Mar. 25, 2019), available at https://www.meritalk.com/articles/japan-pledges-no-ai-killer-robots/.

[14] Alexandra Brzozowski, No progress in UN talks on regulating lethal autonomous weapons, Euractiv (Nov. 22, 2019), available at https://www.euractiv.com/section/global-europe/news/no-progress-in-un-talks-on-regulating-lethal-autonomous-weapons/.

[15] The only notable legislative proposal was the Fundamentally Understanding the Usability and Realistic Evolution of Artificial Intelligence Act of 2017, also known as the FUTURE of Artificial Intelligence Act, which did not aim to regulate AI directly, but instead proposed a Federal Advisory Committee on the Development and Implementation of Artificial Intelligence. The Act has not been re-introduced in the new Congress.

[16] Donald J. Trump, Executive Order on Maintaining American Leadership in Artificial Intelligence, The White House (Feb. 11, 2019), Exec. Order No. 13859, 3 C.F.R. 3967, available at https://www.whitehouse.gov/presidential-actions/executive-order-maintaining-american-leadership-artificial-intelligence/.

[17] Jon Fingas, White House Launches Site to Highlight AI Initiatives, Endgadget (Mar. 20, 2019), available at https://www.engadget.com/2019/03/20/white-house-ai-gov-website/.

[18] For an in-depth analysis, please see our update President Trump Issues Executive Order on “Maintaining American Leadership in Artificial Intelligence.”

[19] The White House, Accelerating America’s Leadership in Artificial Intelligence, Office of Science and Technology Policy (Feb. 11, 2019), available at https://www.whitehouse.gov/briefings-statements/president-donald-j-trump-is-accelerating-americas-leadership-in-artificial-intelligence/.

[20] See, e.g., Jamie Condliffe, In 2017, China is Doubling Down on AI, MIT Technology Review (Jan. 17, 2017), available at https://www.technologyreview.com/s/603378/in-2017-china-is-doubling-down-on-ai/; Cade Metz, As China Marches Forward on A.I., the White House Is Silent, N.Y. Times (Feb. 12, 2018), available at https://www.nytimes.com/2018/02/12/technology/china-trump-artificial-intelligence.html?module=inline.

[21] Supra note 16, section 2(a) (directing federal agencies to prioritize AI investments in their ‘R&D missions’ to encourage ‘sustained investment in AI R&D in collaboration with industry, academia, international partners and allies, and other non-Federal entities to generate technological breakthroughs in AI and related technologies and to rapidly transition those breakthroughs into capabilities that contribute to our economic and national security.’).

[22] Id., section 5 (stating that ‘[h]eads of all agencies shall review their Federal data and models to identify opportunities to increase access and use by the greater non-Federal AI research community in a manner that benefits that community, while protecting safety, security, privacy, and confidentiality’).

[23] The EO asks federal agencies to prioritize fellowship and training programs to prepare for changes relating to AI technologies and promoting science, technology, engineering and mathematics (STEM) education.

[24] In addition, the EO encourages federal agencies to work with other nations in AI development, but also to safeguard the country’s AI resources against adversaries.

[25] NIST’s indirect participation in the development of AI-related standards through the International Organization for Standardization (ISO) may prove to be an early bellwether for future developments.

[26] NIST, U.S. Leadership in AI: a Plan For Federal Engagement in Developing Technical Standards and Related Tools – Draft For Public Comment (July 2, 2019), available .

[27] For instance, the EO established an internal deadline for agencies to submit responsive plans and memoranda for 10 August 2019.

[28] Donald J. Trump, Artificial Intelligence for the American People, the White House (2019), available at https://www.whitehouse.gov/ai/; see also Khari Johnson, The White House Launches ai.gov, VentureBeat (Mar. 19, 2019), available at https://venturebeat.com/2019/03/19/the-white-house-launches-ai-gov/.

[29] Id.; see further our previous legal updates for more details on some of these initiatives: https://www.gibsondunn.com/?search=news&s=&practice%5B%5D=36270.

[30] H.R. 2022, 116th Cong (2019). See https://www.congress.gov/bill/116th-congress/house-bill/2202 or https://lipinski.house.gov/press-releases/lipinski-introduces-bipartisan-legislation-to-bolster-us-leadership-in-ai-research.

[31] H.R. 2022, 116th Cong. (2019). For more details, see https://www.congress.gov/bill/116th-congress/house-bill/2202 or https://lipinski.house.gov/press-releases/lipinski-introduces-bipartisan-legislation-to-bolster-us-leadership-in-ai-research/.

[32] S. 1558, 116th Cong (2019–2020).

[33] The bill also establishes the National AI Research and Development Initiative to identify and minimize ‘inappropriate bias and data sets algorithms’. The requirement for NIST to identify metrics used to establish standards for evaluating AI algorithms and their effectiveness, as well as the quality of training data sets, may be of particular interest to businesses. Moreover, the bill requires the Department of Energy to create an AI research program, building state-of-the-art computing facilities that will be made available to private sector users on a cost-recovery basis.

[34] Press Release, Senator Martin Heinrich, Heinrich, Portman, Schatz Propose National Strategy For Artificial Intelligence; Call For $2.2 Billion Investment In Education, Research & Development (May 21, 2019), available at https://www.heinrich.senate.gov/press-releases/heinrich-portman-schatz-propose-national-strategy-for-artificial-intelligence-call-for-22-billion-investment-in-education-research-and-development.

[35] H.R. 2575, 116th Cong. (2019-2020); S. 3502 – AI in Government Act of 2018, 115th Cong. (2017-2018).

[36] Press Release, Senator Brian Schatz, Schatz, Gardner Introduce Legislation To Improve Federal Government’s Use Of Artificial Intelligence (September 2019), available at https://www.schatz.senate.gov/press-releases/schatz-gardner-introduce-legislation-to-improve-federal-governments-use-of-artificial-intelligence; see also Tajha Chappellet-Lanier, Artificial Intelligence in Government Act is back, with ‘smart and effective’ use on senators’ minds (May 8, 2019), available at https://www.fedscoop.com/artificial-intelligence-in-government-act-returns.

[37] Note that, in companion bills SB-5527 and HB-1655, introduced on January 23, 2019, Washington State lawmakers drafted a comprehensive piece of legislation aimed at governing the use of automated decision systems by state agencies, including the use of automated decision-making in the triggering of automated weapon systems. In addition to addressing the fact that eliminating algorithmic-based bias requires consideration of fairness, accountability, and transparency, the bills also include a private right of action. According to the bills’ sponsors, automated decision systems are rapidly being adopted to make or assist in core decisions in a variety of government and business functions, including criminal justice, health care, education, employment, public benefits, insurance, and commerce, and are often unregulated and deployed without public knowledge. Under the new law, in using an automated decision system, an agency would be prohibited from discriminating against an individual, or treating an individual less favorably than another on the basis of one or more of a list of factors such as race, national origin, sex, or age. The bills were reintroduced in the 2020 session and remain in Committee. SB 5527, Reg. Sess. 2019-2020 (Wash. 2020); HB-1655, Reg. Sess. 2010-2020 (Wash. 2020).

[38] White House AI Order Emphasizes Use for Citizen Experience, Meritalk (Apr. 18, 2019), available at https://www.meritalk.com/articles/white-house-ai-order-emphasizes-use-for-citizen-experience/.

[39] Director of the Office of Management and Budget, Guidance for Regulation of Artificial Intelligence Applications (Jan. 7, 2020), available at https://www.whitehouse.gov/wp-content/uploads/2020/01/Draft-OMB-Memo-on-Regulation-of-AI-1-7-19.pdf.

[40] Id., at 2-3.

[41] H.R. 5515, 115th Cong. (2019). See https://www.congress.gov/bill/115th-congress/house-bill/5515/text.

[42] See Cronk, Terri Moon, DoD Unveils Its Artificial Intelligence Strategy (February 12, 2019) at https://www.defense.gov/Newsroom/News/Article/Article/1755942/dod-unveils-its-artificial-intelligence-strategy/. In particular, the JAIC director’s duties include, among other things, developing plans for the adoption of artificial intelligence technologies by the military and working with private companies, universities and nonprofit research institutions toward that end.

[43] Summary of the 2019 Department of Defense Artificial Intelligence Strategy, Harnessing AI to Advance Our Security and Prosperity (https://media.defense.gov/2019/Feb/12/2002088963/-1/-1/1/SUMMARY-OF-DOD-AISTRATEGY.PDF).

[44] Another potentially significant effort is the work currently being performed under the direction of DARPA on developing explainable AI systems. See https://www.darpa.mil/program/explainable-artificial-intelligence. Because it can be difficult to understand exactly how a machine learning algorithm arrives at a particular conclusion or decision, some have referred to artificial intelligence as being a ‘black box’ that is opaque in its reasoning. However, a black box is not always an acceptable operating paradigm, particularly in the context of battlefield decisions, within which it will be important for human operators of AI-driven systems to understand why particular decisions are being made to ensure trust and appropriate oversight of critical decisions. As a result, DARPA has been encouraging the development of new technologies to explain and improve machine–human understanding and interaction. See also DARPA’s ‘AI Next Campaign’ (https://www.darpa.mil/work-with-us/ai-next-campaign).

[45] Id. at 9. See also id. at 15 (the JAIC ‘will articulate its vision and guiding principles for using AI in a lawful and ethical manner to promote our values’.); in addition, under the 2019 NDAA, one duty of the JAIC director is to develop legal and ethical guidelines for the use of AI systems. https://www.govinfo.gov/ content/pkg/BILLS-115hr5515enr/pdf/BILLS-115hr5515enr.pdf

[46] Calls for bans or at least limits on so-called ‘killer robots’ go back several years, and even provoked several thousand signatories, including many leading AI researchers, to the Future of Life Institute’s pledge. See https://futureoflife.org/lethal-autonomous-weapons-pledge.

[47] Allocations across the National Science Foundation, the Department of Energy’s Office of Science and the Defense Advanced Research Projects Agency and the Department of Defense’s Joint AI Center reach a combined $1.724 billion — with portions of an additional $150 million allocation for the Department of Agriculture and the National Institutes of Health going to AI research. Note that the draft budget also proposes a cut of 19%, or $154 million, to NIST’s $653 million budget. See https://www.sciencemag.org/news/2020/02/trump-s-2021-budget-drowns-science-agencies-red-ink-again.

[48] Such as the U.S. Constitution, international treaties and the Pentagon’s Law of War.

[49] Defense Innovation Board, AI Principles: Recommendations on the Ethical Use of Artificial Intelligence by the Department of Defense (Oct. 31, 2019), available at https://media.defense.gov/2019/Oct/31/2002204458/
-1/-1/0/DIB_AI_PRINCIPLES_PRIMARY_DOCUMENT.PDF.

[50] Jack Corrigan, Defense Innovation Board Lays Out 5 Key Principles for Ethical AI, Nextgov (Oct. 31, 2019), available at https://www.nextgov.com/emerging-tech/2019/10/defense-innovation-board-lays-out-5-key-principles-ethical-ai/161008/.

[51] Daniel Wilson, New Ethics Framework May Draw AI Firms To DOD, Law360 (Nov. 8, 2019), available at https://www.law360.com/articles/1217965/new-ethics-framework-may-draw-ai-firms-to-dod.

[52] Exec. Office of the U.S. President, The National Artificial Intelligence Research and Development Strategic Plan: 2019 Update (June 2019), available at https://www.whitehouse.gov/wp-content/uploads/2019/06/National-AI-Research-and-Development-Strategic-Plan-2019-Update-June-2019.pdf, The updated plan also highlights what progress federal agencies have made with respect to the original seven focus areas: make long-term investments in AI research; develop effective methods for human-AI collaboration; understand and address the ethical, legal and societal implications of AI; ensure the safety and security of AI systems; develop shared public datasets and environments for AI training and testing; measure and evaluate AI technologies through standards and benchmarks; and better understand the national AI R&D workforce needs.

[53] Id., at 42.

[54] National Security Commission on Artificial Intelligence, Interim Report (Nov. 2019), available at https://www.epic.org/foia/epic-v-ai-commission/AI-Commission-Interim-Report-Nov-2019.pdf

[55] Id., at 22.

[56] Id., at 25.

[57] Id., at 31.

[58] Id., at 36.

[59] Id., at 18.

[60] Id., at 41&44.

[61] U.S. Department of Commerce, Press Release, U.S. Department of Commerce Adds 28 Chinese Organizations to its Entity List (Oct. 7, 2019), available at https://www.commerce.gov/news/press-releases/2019/10/us-department-commerce-adds-28-chinese-organizations-its-entity-list.

[62] Anna Swanson and Paul Mozur, U.S. Blacklists 28 Chinese Entities Over Abuses in Xinjiang, N.Y. Times (Oct. 7, 2019), available at https://www.nytimes.com/2019/10/07/us/politics/us-to-blacklist-28-chinese-entities-over-abuses-in-xinjiang.html.

[63] U.S. Federal Register, Addition of Software Specially Designed To Automate the Analysis of Geospatial Imagery to the Export Control Classification Number 0Y521 Series (Docket No. BIS-2019-0031) (Jan. 6, 2020), available at https://www.federalregister.gov/documents/2020/01/06/2019-27649/addition-of-software-specially-designed-to-automate-the-analysis-of-geospatial-imagery-to-the-export.

[64] The European Commission (EC) enacted a proposal titled: ‘The Communication From the Commission to the European Parliament, the European Council, the European Economic and Social Committee, and the Committee of the Regions: Artificial Intelligence for Europe’ (25 April 2018), https://ec.europa.eu/digital-single-market/en/news/communication-artificial-intelligence-europe. The Communication set out the following regulatory proposals for AI: calls for new funding, pledges for investment in explainable AI ‘beyond 2020’, plans for evaluation of AI regulation, proposes that the Commission will support the use of AI in the justice system, pledges to draft AI ethics guidelines by the end of the year, proposes dedicated retraining schemes, and calls for prompt adoption of the proposed ePrivacy Regulation. Likewise, an April 2018 UK Select Committee Report on AI encouraged the UK government to establish a national AI strategy and proposed an ‘AI Code’ with five principles, emphasizing ideals such as fairness and developing for the common good – mirroring the EU’s AI Ethics Guidelines. ‘AI Policy – United Kingdom,’ available at https://futureoflife.org/ai-policy-united-kingdom/?cn-reloaded=1.

[65] H. Mark Lyon, Gearing Up For The EU’s Next Regulatory Push: AI, LA & SF Daily Journal (Oct. 11, 2019), available at https://www.gibsondunn.com/wp-content/uploads/2019/10/Lyon-Gearing-up-for-the-EUs-next-regulatory-push-AI-Daily-Journal-10-11-2019.pdf.

[66] See further, Ahmed Baladi, Can GDPR hinder AI made in Europe?, Cybersecurity Law Report (July 10, 2019), available at https://www.gibsondunn.com/can-gdpr-hinder-ai-made-in-europe/.

[67] Ursula von der Leyen, A Union that strives for more: My agenda for Europe, available at https://ec.europa.eu/commission/sites/beta-political/files/political-guidelines-next-commission_en.pdf. Note that the von der Leyen commission was slated to begin on November 1, 2019, but due to problems with filling three of the commissioners’ seats, it was delayed until December 1, 2019 (thus pushing back the 100-day deadline).

[68] Oscar Williams, New European Commission president pledges GDPR-style AI legislation, New Statesman (Nov. 28, 2019), available at https://tech.newstatesman.com/policy/ursula-von-der-leyen-ai-legislation.

[69] Reuters, EU Drops Idea of Facial Recognition Ban in Public Areas, N.Y. Times (Jan. 30, 2020), available at https://www.nytimes.com/reuters/2020/01/30/technology/29reuters-eu-ai.html.

[70] European Commission, Structure for the white paper on Artificial Intelligence, Euractiv (Jan. 2, 2020), available at https://www.euractiv.com/wp-content/uploads/sites/2/2020/01/AI-white-paper-EURACTIV.pdf.

[71] See further, H. Mark Lyon, Gearing up for the EU’s next regulatory push: AI, LA & SF Daily Journal (Oct. 11, 2019), available at https://www.gibsondunn.com/wp-content/uploads/2019/10/Lyon-Gearing-up-for-the-EUs-next-regulatory-push-AI-Daily-Journal-10-11-2019.pdf.

[72] EC, Artificial Intelligence for Europe, COM(2018) 237 (Apr. 25, 2018), available at https://ec.europa.eu/digital-single-market/en/news/communication-artificial-intelligence-europe.

[73] AI HLEG, Ethics Guidelines for Trustworthy AI, Guidelines (Apr. 8, 2019), available at https://ec.europa.eu/newsroom/dae/document.cfm?doc_id=60419.

[74] In a speech at the European Parliament on November 27, 2019, von der Leyen said that she was in favor of AI-focused legislation similar to the GDPR. The Commission is also likely to draw on the work of its high-level expert group on AI, which outlined a series of principles earlier this year aimed at ensuring companies deploy artificial intelligence in a way that is fair, safe and accountable. In a keynote speech at the World Economic Forum on January 22, 2020, von der Leyen stated that the GDPR had already set a pattern for the world and that the EU would “have to set a similar frame for artificial intelligence, too.”

[75] AI HLEG, Trustworthy AI Assessment List, List (Apr. 8, 2019), available at https://ec.europa.eu/newsroom/dae/document.cfm?doc_id=60440.

[76] EC, Pilot the Assessment List of the Ethics Guidelines for Trustworthy AI, website article available at https://ec.europa.eu/futurium/en/ethics-guidelines-trustworthy-ai/register-piloting-process-0.

[77] Id.

[78] Artificial Intelligence: Commission takes forward its work on ethical guidelines, Press Release, Apr. 8, 2019, available at http://europa.eu/rapid/press-release_IP-19-1893_en.htm.

[79] German Federal Ministry for Justice and Consumer Protection, Opinion of the Data Ethics Commission, Executive Summary (October 2019), available at http://bit.ly/373RGqI.

[80] Jeremy Feigelson, Jim Pastore, Anna Gressel and Friedrich Popp, German Report May Be Road Map For Future AI Regulation, Law360 (Nov. 12, 2019), available at https://www.law360.com/articles/1218560/german-report-may-be-road-map-for-future-ai-regulation.

[81] German Federal Ministry for Justice and Consumer Protection, Opinion of the Data Ethics Commission, supra, note 33 at 7.

[82] Id., at 19-20.

[83] Id., at 10.

[84] Id., at 26.

[85] David Meyer, A.I. Regulation Is Coming Soon. Here’s What the Future May Hold, Fortune (Oct. 24, 2019), available at https://fortune.com/2019/10/24/german-eu-data-ethics-ai-regulation/.

[86] German Federal Ministry for Justice and Consumer Protection, Opinion of the Data Ethics Commission, supra, note 33 at 5.

[87] See also Kalev Leetaru, Why Do We Fix AI Bias But Ignore Accessibility Bias?, Forbes (July 6, 2019), available at https://www.forbes.com/sites/kalevleetaru/2019/07/06/why-do-we-fix-ai-bias-but-ignore-accessibility-bias/#55e7c777902d; Alina Tugend, Exposing the Bias Embedded in Tech, N.Y. Times (June 17, 2019), available at https://www.nytimes.com/2019/06/17/business/artificial-intelligence-bias-tech.html.

[88] Jake Silberg & James Manyika, Tackling Bias in Artificial Intelligence (and in Humans), McKinsey Global Institute (June 2019), available at https://www.mckinsey.com/featured-insights/artificial-intelligence/tackling-bias-in-artificial-intelligence-and-in-humans.

[89] Nicol Turner Lee, Paul Resnick & Genie Barton, Algorithmic Bias Detection and Mitigation: Best Practices and Policies to Reduce Consumer Harms, Brookings Institute (May 22, 2019), available at https://www.brookings.edu/research/algorithmic-bias-detection-and-mitigation-best-practices-and-policies-to-reduce-consumer-harms/.

[90] See also the French government’s recent law, encoded in Article 33 of the Justice Reform Act, prohibiting anyone—especially legal tech companies focused on litigation prediction and analytics—from publicly revealing the pattern of judges’ behavior in relation to court decisions, France Bans Judge Analytics, 5 Years In Prison For Rule Breakers, Artificial Lawyer (June 4, 2019), available at https://www.artificiallawyer.com/2019/06/04/france-bans-judge-analytics-5-years-in-prison-for-rule-breakers/.

[91] See, e.g., Karen Hao, Congress Wants To Protect You From Biased Algorithms, Deepfakes, And Other Bad AI, MIT Review (15 April 2019), available at https://www.technologyreview.com/s/613310/congress-wantsto- protect-you-from-biased-algorithms-deepfakes-and-other-bad-ai/; Meredith Whittaker, et al, AI Now Report 2018, AI Now Institute, 2.2.1 (December 2018), available at https://ainowinstitute.org/AI_Now_2018_Report. pdf; Russell Brandom, Congress Thinks Google Has a Bias Problem—Does It?, The Verge (12 December 2018), available at https://www.theverge.com/2018/12/12/18136619/google-bias-sundar-pichai-google-hearing.

[92] H.R. Res. 153, 116th Cong. (1st Sess. 2019).

[93] Assemb. Con. Res. 215, Reg. Sess. 2018-2019 (Cal. 2018) (enacted) (expressing the support of the legislature for the “Asilomar AI Principles”—a set of 23 principles developed through a collaboration between AI researchers, economists, legal scholars, ethicists and philosophers that met in Asilomar, California in January 2017 and categorized into “research issues,” “ethics and values,” and “longer-term issues” designed to promote the safe and beneficial development of AI—as “guiding values for the development of artificial intelligence and of related public policy”).

[94] OECD Principles on AI (May 22, 2019) (stating that AI systems should benefit people, be inclusive, transparent, and safe, and their creators should be accountable), available at http://www.oecd.org/going-digital/ai/principles/.

[95] Press Release, Cory Booker, Booker, Wyden, Clarke Introduce Bill Requiring Companies To Target Bias In Corporate Algorithms (Apr. 10, 2019), available at https://www.booker.senate.gov/?p=press_release&id=903; see also S. Res. __, 116th Cong. (2019).

[96] H.R. Res. 2231, 116th Cong. (1st Sess. 2019).

[97] Supra, note 66.

[98] Id.

[99] See Byungkwon Lim et al., A Glimpse into the Potential Future of AI Regulation, Law360 (April 10, 2019), available at https://www.law360.com/articles/1158677/a-glimpse-into-the-potential-future-of-ai-regulation.

[100] S.3127 – Bot Disclosure and Accountability Act of 2018, 115th Cong (2018), available at https://www.congress.gov/bill/115th-congress/senate-bill/3127 and S.2125 Bot Disclosure and Accountability Act of 2019, 116th Cong (2019), available at https://www.congress.gov/bill/116th-congress/senate-bill/2125.

[101] SB 1001, Bolstering Online Transparency Act (CA 2017), available at https://leginfo.legislature.ca.gov/faces/

billTextClient.xhtml?bill_id=201720180SB1001. We previously provided a detailed analysis of the new law in our client alert New California Security of Connected Devices Law and CCPA Amendments.

[102] Filter Bubble Transparency Act, S. 2763, 116th Cong. (2019). The bill’s sponsors are Senators Marsha Blackburn (R-Tenn.), John Thune (R-S.D.), Richard Blumenthal (D-Conn.), Jerry Moran (R-Kan.)—all members of the Senate Committee on Commerce, Science, and Transportation, which has jurisdiction over the internet and consumer protection—and Mark Warner (D-Va.).

[103] Blackburn Joins Thune on Bipartisan Bill to Increase Internet Platform Transparency & Provide Consumers with Greater Control Over Digital Content, Marsha Blackburn, U.S. Senator for Tennessee (Oct. 31, 2019), https://www.blackburn.senate.gov/blackburn-joins-thune-bipartisan-bill-increase-internet-platform-transparency-provide-consumers.

[104] Supra, note 74; see also Zoe Schiffer, ‘Filter Bubble’ author Eli Pariser on why we need publicly owned social networks, The Verge (Nov. 12, 2019), available at https://www.theverge.com/2019/11/5/20943634/senate-filter-bubble-transparency-act-algorithm-personalization-targeting-bill.

[105] Filter Bubble Transparency Act, supra n.1, at 2(4)(A)–(B). The bill provides that it is also applicable to common carriers that are subject to the Communications Act of 1934 and to “organizations not organized to carry on business for their own profit or that of their members.” Id. at 4(B)(3).

[106] Id. at 2(B). The term “algorithmic ranking system” is broadly defined and encompasses any computational process—including “one derived from algorithmic decision-making, machine learning, statistical analysis, or other data processing or artificial intelligence techniques”—that is used to determine the order in which a set of information is provided to a user on a covered internet platform. Examples include “the ranking of search results, the provision of content recommendations, the display of social media posts, or any other method of automated content selection.”

[107] Filter Bubble Transparency Act, supra n.1, at 2(1).

[108] See id. at 2(5)(B).

[109] Id. at 5(A), (C).

[110] Id. at 5(A).

[111] Id. at 3(A)–(B) (emphasis added).

[112] Supra, note 74.

[113] Adi Robertson, The Senate’s secret algorithms bill doesn’t actually fight secret algorithms, The Verge (Nov. 5, 2019), available at https://www.theverge.com/2019/11/5/20943634/senate-filter-bubble-transparency-act-algorithm-personalization-targeting-bill.

[114] The bill also exempts platforms that are operated for the sole purpose of conducting research that is not made for direct or indirect profit. Id. at 2(4)(A)–(B). Moreover, the bill does not cover contractors and subcontractors that receive rights to access indexes of web pages on the internet for the purpose of operating an internet search engine (i.e., downstream providers) from the respective upstream providers if “the search engine is operated by a downstream provider with fewer than 1,000 employees” and “the search engine uses an index of web pages on the internet to which such provider received access under a search syndication contract.” Id. at 3(B)(2).

[115] On May 20, 2019, New Jersey introduced a similar bill, New Jersey Algorithmic Accountability Act (A.B. 5430), which requires covered entities to conduct impact assessments on “high-risk” automated decisions systems and information systems. New Jersey Algorithmic Accountability Act, A.B. 5430, 218th Leg., 2019 Reg. Sess. (N.J. 2019).

[116] U.S. H.R. Comm. on Oversight and Reform, Facial Recognition Technology (Part II): Ensuring Transparency in Government Use (June 4, 2019), available at https://oversight.house.gov/legislation/hearings/facial-recognition-technology-part-ii-ensuring-transparency-in-government-use.

[117] S. 847, 116th Cong. (1st Sess. 2019).

[118] H.R. 3875, 116th Cong. (2019).

[119] H.R. 4008, 116th Cong. (2019).

[120] FACE Protection Act of 2019, H.R. 4021, 116th Cong. (2019).

[121] U.S. H.R. Comm. on Oversight and Reform, Facial Recognition Technology (Part II): Ensuring Transparency in Government Use (June 4, 2019), available at https://oversight.house.gov/legislation/hearings/facial-recognition-technology-part-ii-ensuring-transparency-in-government-use.

[122] Khari Johnson, Congress moves towards facial recognition regulation, Venture Beat (Jan. 15, 2020), available at https://venturebeat.com/2020/01/15/congress-moves-toward-facial-recognition-regulation/.

[123] A.B. 1215 2019–2020 Reg. Sess. (Cal. 2019); see also Anita Chabria, California could soon ban facial recognition technology on police body cameras (Sept. 12, 2019), available at https://www.latimes.com/california/story/2019-09-12/facial-recognition-police-body-cameras-california-legislation.

[124] A.B. 1215 2019-2020 Reg. Sess. (Cal. 2019).

[125] Id., at 44.

[126] Id., at 1(c).

[127] See San Francisco Ordinance No. 103-19, the ‘Stop Secret Surveillance’ ordinance, effective 31 May 2019 (banning the use of facial recognition software by public departments within San Francisco, California); Somerville Ordinance No. 2019-16, the ‘Face Surveillance Full Ban Ordinance’, effective 27 June 2019 (banning use of facial recognition by the City of Somerville, Massachusetts or any of its officials); Oakland Ordinance No. 18-1891, ‘Ordinance Amending Oakland Municipal Code Chapter 9.65 to Prohibit the City of Oakland from Acquiring and/or Using Real-Time Face Recognition Technology’, preliminary approval 16 July 2019, final approval 17 September 2019 (bans use by city of Oakland, California and public officials of real-time facial recognition); Proposed Amendment attached to Cambridge Policy Order POR 2019 #255, approved on 30 July 2019 for review by Public Safety Committee (proposing ban on use of facial recognition technology by City of Cambridge, Massachusetts or any City staff); Attachment 5 to Berkeley Action Calendar for 11 June 2019, ‘Amending Berkeley Municipal Code Chapter 2.99 to Prohibit City Use of Face Recognition Technology’, voted for review by Public Safety Committee on 11 June 2019 and voted for continued review by Public Safety Committee on 17 July 2019 (proposing ban on use of facial recognition technology by staff and City of Berkeley, California). All of these ordinances incorporated an outright ban of use of facial recognition technology, regardless of the actual form or application of such technology. For a view on how such a reactionary ban is an inappropriate way to regulate AI technologies, see Lyon, H Mark, ‘Before We Regulate’, Daily Journal (26 June 2019) available at https://www.gibsondunn.com/before-we-regulate.

[128] Sarah Ravani, Oakland bans use of facial recognition technology, citing bias concerns, SF Chronicle (July 17, 2019), available at https://www.sfchronicle.com/bayarea/article/Oakland-bans-use-of-facial-recognition-14101253.php; see also, Cade Metz, Facial Recognition Tech Is Growing Stronger, Thanks to Your Face, N.Y. Times (July 13, 2019), available at https://www.nytimes.com/2019/07/13/technology/databases-faces-facial-recognition-technology.html.

[129] Levi Sumagaysay, Berkeley bans facial recognition, Mercury News (Oct. 16, 2019), available at https://www.mercurynews.com/2019/10/16/berkeley-bans-facial-recognition/.

[130] AB 730, AB 602 (California); SB 751 (Texas); HB 2678 (Virginia); HR 3230, 116th Congress (U.S. House of Representatives); S 2065, 116th Congress (U.S. Senate).

[131] Letter from Adam Schiff, U.S. Representative, Stephanie Murphy, U.S. Representative & Carlos Curbelo, U.S. Representative to Hon. Daniel R. Coats, Dir. of Nat’l Intelligence (Sept. 13, 2018).

[132] Id.

[133] S. 2065, 116th Congress (U.S. Senate).

[134] H.R. 3600, 116th Congress (U.S. House of Representatives).

[135] H.R. 3230, 116th Congress (U.S. House of Representatives).

[136] H.R. 4355, 116th Congress (U.S. House of Representatives).

[137] A.B. 730 2019–2020 Reg. Sess. (Cal. 2019).

[138] A.B. 602 2019-2020 Reg. Sess. (Cal. 2019).

[139] H.R. 3388, 115th Cong. (2017).

[140] U.S. Senate Committee on Commerce, Science and Transportation, Press Release, Oct. 24, 2017, available at https://www.commerce.senate.gov/public/index.cfm/pressreleases?ID=BA5E2D29-2BF3-4FC7-A79D-58B9E186412C.

[141] Letter from Democratic Senators to U.S. Senate Committee on Commerce, Science and Transportation (Mar. 14, 2018), available at https://morningconsult.com/wp-content/uploads/2018/11/2018.03.14-AV-START-Act-letter.pdf.

[142] U.S. Senate Committee on Commerce, Science and Transportation, Press Release, Oct. 24, 2017, available at https://www.commerce.senate.gov/public/index.cfm/pressreleases?ID=BA5E2D29-2BF3-4FC7-A79D-58B9E186412C.

[143] Makena Kelly, Congress wants the self-driving car industry’s help to draft a new AV bill, The Verge (July 31, 2019), available at https://www.theverge.com/2019/7/31/20748582/congress-self-driving-cars-bill-energy-commerce-senate-regulation.

[144] U.S. Dep’t of Transp., Ensuring American Leadership in Automated Vehicle Technologies: Automated Vehicles 4.0 (Jan. 2020), available at https://www.transportation.gov/sites/dot.gov/files/docs/policy-initiatives/automated-vehicles/360956/ensuringamericanleadershipav4.pdf.

[145] U.S. Dep’t of Transp., Preparing for the Future of Transportation: Automated Vehicles 3.0 (Sept. 2017), available at https://www.transportation.gov/sites/dot.gov/files/docs/policy-initiatives/automated-vehicles/320711/preparing-future-transportation-automated-vehicle-30.pdf; see further our Artificial Intelligence and Autonomous Systems Legal Update (4Q18).

[146] General Motors, LLC-Receipt of Petition for Temporary Exemption from Various Requirements of the Safety Standards for an All Electric Vehicle with an Automated Driving System, 84 Fed. Reg. 10182.

[147] Docket No. NHTSA-2019-0036, ‘Removing Regulatory Barriers for Vehicles With Automated Driving Systems’, 84 Fed Reg 24,433 (28 May 2019) (to be codified at 49 CFR 571); see also ‘Removing Regulatory Barriers for Vehicles with Automated Driving Systems’, 83 Fed Reg 2607, 2607 (proposed 5 March 2018) (to be codified at 49 CFR 571). Thus far, the comments submitted generally support GM’s petition for temporary exemption and the removal of regulatory barriers to the compliance certification of ADS-DVs. Some commentators have raised concerns that there is insufficient information in the petition to establish safety equivalence between traditionally operated vehicles and ADS-DVs, and regarding the ability of ADS-DVs to safely operate in unexpected and emergency situations. However, it is likely that NHTSA will grant petitions for temporary exemption to facilitate the development of ADS technology, contingent on extensive data-sharing requirements and a narrow geographic scope of operation. In addition, the Federal Motor Carrier Safety Administration also issued a request for comments on proposed rule-making for Federal Motor Carrier Safety Regulations that may need to be reconsidered for Automated Driving System- Dedicated Vehicles (ADS-DVs). Docket No. FMCSA-2018-0037. Safe Integration of Automated Driving Systems-Equipped Commercial Motor Vehicles, 84 Fed Reg 24,449 (28 May 2019).

[148] Nat’l Conference of State Legislatures, Autonomous Vehicles State Bill Tracking Database (Jan. 5, 2020), available at http://www.ncsl.org/research/transportation/autonomous-vehicles-legislative-database.aspx.

[149] Dan Robitzki, Florida Law Would Allow Self-Driving Cars With No Safety Drivers, Futurism (Jan. 29, 2019) available at https://futurism.com/florida-law-self-driving-cars.

[150] Andrew Glambrone, Self-Driving Cars Are Coming. D.C. Lawmakers Want To Regulate Them, Curbed (Apr. 3, 2019), available at https://dc.curbed.com/2019/4/3/18294167/autonomous-vehicles-dc-self-driving-cars-regulations.

[151] Id., see further the Autonomous Vehicles Testing Program Amendment Act of 2019, available at http://lims.dccouncil.us/Download/42211/B23-0232-Introduction.pdf.

[152] State of California Department of Motor Vehicles, Autonomous Light-Duty Motor Trucks (Delivery Vehicles), available at https://www.dmv.ca.gov/portal/dmv/detail/vr/autonomous/bkgd.

[153] S.B. 59, 2019–2020 Reg. Sess. (Cal. 2019).

[154] S.B. 336, 2019–2020 Reg. Sess. (Cal. 2019).

[155] Governor Ron DeSantis Signs CS/HB 311: Autonomous Vehicles (June 13, 2019), available at https://www.flgov.com/2019/06/13/governor-ron-desantis-signs-cs-hb-311-autonomous-vehicles/.

[156] S.B. 365, 57th Leg., Reg. Sess. (Okla. 2019).

[157] H.B. 1078, 2019–2020 Reg. Sess. (Pa. 2019).

[158] In addition to the legislation referenced in this section, and as we discuss in more detail in our U.S. Cybersecurity and Data Privacy Outlook and Review – 2020, litigation also continued around Illinois’ Biometric Information Privacy Act (“BIPA”), including litigation predicated on the use of facial recognition technology.

[159] See S. 2398, 116th Congress (Senate).

[160] For more information, see our prior client alert, California Consumer Privacy Act: 2019 Final Amendments Signed, available at https://www.gibsondunn.com/california-consumer-privacy-act-2019-final-amendments-signed/.

[161] Again, for more information on the proposed regulations for CCPA, please see our prior client alert, California Consumer Privacy Act Update: Regulatory Update, available at https://www.gibsondunn.com/california-consumer-privacy-act-update-regulatory-update/.

[162] A.B. 1395, 2019–2010 Reg. Sess. (Cal. 2019).

[163] Request for Comments on Patenting Artificial Intelligence Inventions, 84 Fed. Reg. 44889, 44889 (Aug. 27, 2019); see also our client alert USPTO Requests Public Comments on Patenting Artificial Intelligence Inventions.

[164] See further Mark Lyon, Alison Watkins and Ryan Iwahashi, When AI Creates IP: Inventorship Issues To Consider, Law360 (Aug. 10, 2017), available at https://www.law360.com/articles/950313?scroll=1&related=1.

[165] Ryan Davis, Law Shouldn’t Let AI Be An Inventor On Patents, USPTO Told, Law360 (Nov. 13, 2019), available at https://www.law360.com/articles/1218939/law-shouldn-t-let-ai-be-an-inventor-on-patents-uspto-told.

[166] WIPO Begins Public Consultation Process on Artificial Intelligence and Intellectual Property Policy, Press Release (Dec. 13, 2019), available at https://www.wipo.int/pressroom/en/articles/2019/article_0017.html.

[167] H.R. 4368, 116th Congress (U.S. House of Representatives).

[168] Press Release, Rep. Takano Introduces the Justice in Forensic Algorithms Act to Protect Defendants’ Due Process Rights in the Criminal Justice System (Sept. 17, 2019), available at https://takano.house.gov/newsroom/press-releases/rep-takano-introduces-the-justice-in-forensic-algorithms-act-to-protect-defendants-due-process-rights-in-the-criminal-justice-system.

[169] Karen Hao, AI Is Sending People To Jail – And Getting It Wrong, MIT Technology Review (Jan. 21, 2019), available at https://www.technologyreview.com/s/612775/algorithms-criminal-justice-ai/. See also Rod McCullom, Facial Recognition Technology is Both Biased and Understudied, UnDark (May 17, 2017), available at https://undark.org/article/facial-recognition-technology-biased-understudied/.

[170] See Karen Hao, Police Across the US Are Training Crime-Predicting AIs on Falsified Data, MIT Technology Review (Feb. 13, 2019), available at https://www.technologyreview.com/s/612957/predictive-policing-algorithms-ai-crime-dirty-data/.

[171] Meredith Whittaker, et al, AI Now Report 2018, AI Now Institute, 2.2.1 (December 2018), available at https://ainowinstitute.org/AI_Now_2018_Report; see also Rashida Richardson Schultz, Jason Schultz, and Kate Crawford, Dirty Data, Bad Predictions: How Civil Rights Violations Impact Police Data, Predictive Policing Systems, and Justice (Feb. 13, 2019). New York University Law Review Online, Forthcoming, available at SSRN: https://ssrn.com/abstract=3333423.

[172] Meredith Whittaker, et al, AI Now Report 2018, AI Now Institute, 2.2.1 (December 2018), available at https://ainowinstitute.org/AI_Now_2018_Report

[173] H.B. 2557, 2019-2010 Reg. Sess. (Ill. 2019) (101st Gen. Assembly).

[174] See, e.g., A.B. 1215 2019–2020 Reg. Sess. (Cal. 2019); see also Anita Chabria, California could soon ban facial recognition technology on police body cameras (Sept. 12, 2019), available at https://www.latimes.com/california/story/2019-09-12/facial-recognition-police-body-cameras-california-legislation.

[175] See, e.g., Stop LAPD Spying Coalition v. City of Los Angeles, BS172216 (Cal. Super. Ct. 2018) – currently pending.

[176] For example, AI has been used in robot-assisted surgery in select fields for years, and studies have shown that AI-assisted procedures can result in far fewer complications. Brian Kalis, Matt Collier and Richard Fu, ‘10 Promising AI Applications in Health Care’, Harvard Business Review (10 May 2018), available at https://hbr.org/2018/05/10-promising-ai-applications-in-health-care. Yet, The New York Times published an article in March 2019 warning of healthcare AI’s potential failures, including small changes in vernacular leading to vastly disparate results (eg, ‘alcohol abuse’ leading to a different diagnosis than ‘alcohol dependence’); see Cade Metz and Craig S Smith, ‘Warning of a Dark Side to A.I. in Health Care’, The New York Times (21 March 2019), available at nytimes.com/2019/03/21/science/health-medicine-artificial-intelligence.html. And these issues are backed by studies, including one released by Science – one of the highest acclaimed journals – just prior to the article, which discusses how ‘vulnerabilities allow a small, carefully designed change in how inputs are presented to a system to completely alter its outputs, causing it to confidently arrive at manifestly wrong conclusions.’ Samuel G Finlayson, et al, ‘Adversarial attacks on medical machine learning’, SCIENCE 363:6433, pp. 1287–1289 (22 March 2019) See https://science.sciencemag.org/content/363/6433/1287.

[177] U.S. Food & Drug Administration, Proposed Regulatory Framework for Modifications to Artificial Intelligence/Machine Learning (AI/ML)-Based Software as a Medical Device (SaMD), at 2 (2 April 2019), available at https://www.fda.gov/media/122535/download.

[178] The paper mentions that AI-based SaMDs have been approved by the FDA, but they are generally ‘locked’ algorithms, and any changes would be expected to go through pre-market review. This proposal attempts to anticipate continuously-adapting AI-based SaMD products.

[179] Katie Grzechnik Neill, Rep. Waters Announces Task Forces on Fintech and Artificial Intelligence (May 13, 2019), available at https://www.insidearm.com/news/00045030-rep-waters-announces-all-democrat-task-fo.

[180] See Scott Likens, How Artificial Intelligence Is Already Disrupting Financial Services, Barrons (May 16, 2019), available at https://www.barrons.com/articles/how-artificial-intelligence-is-already-disrupting-financial-services-51558008001.

[181] H.R. 4476, 116th Congress (U.S. House of Representatives).

[182] Id. (including the Securities and Exchange Commission, Commodity Futures Trading Commission, Federal Deposit Insurance Corp., Federal Reserve, Office of the Comptroller of the Currency, the Consumer Financial Protection Bureau, the National Credit Union Association and the Federal Housing Finance Agency).

[183] Id.

[184] Robert Booth, Unilever saves on recruiters by using AI to assess job interviews, The Guardian (Oct. 25, 2019), available at https://www.theguardian.com/technology/2019/oct/25/unilever-saves-on-recruiters-by-using-ai-to-assess-job-interviews; Lloyd Chinn & Thomas Fiascone, AI In Hiring: Legislative Responses And Litigation Potential, Law360 (Nov. 25, 2019), available at https://www.law360.com/illinois/articles/1220318/ai-in-hiring-legislative-responses-and-litigation-potential.

[185] H.R. 827 – AI JOBS Act of 2019, 116th Cong (2019), available at https://www.congress.gov/bill/116thcongress/house-bill/827/text.

[186] S. 2468, 116th Congress (U.S. Senate).

[187] Id.

[188] Drew Harwell, A face-scanning algorithm increasingly decides whether you deserve the job, Wash. Post (Oct. 25, 2019), available at https://www.washingtonpost.com/technology/2019/10/22/ai-hiring-face-scanning-algorithm-increasingly-decides-whether-you-deserve-job/.

[189] H.B. 2557, 2019-2020 Reg. Sess. (Ill. 2019) (101st Gen. Assembly), available at http://www.ilga.gov/legislation/101/HB/PDF/10100HB2557lv.pdf.

[190] Drew Harwell, A face-scanning algorithm increasingly decides whether you deserve the job, Wash. Post (Oct. 25, 2019), available at https://www.washingtonpost.com/technology/2019/10/22/ai-hiring-face-scanning-algorithm-increasingly-decides-whether-you-deserve-job/.

[191] For example, Washington State introduced legislation on January 23, 2019 that provides a private right of action and seeks to address the elimination of algorithmic bias through a careful consideration of fairness, accountability, and transparency. See Brian Higgins, Washington State Seeks to Root Out Bias in Artificial Intelligence Systems, Artificial Intelligence Technology and the Law (Feb. 6, 2019), available at http://aitechnologylaw.com/2019/02/washington-state-seeks-to-root-out-bias-in-artificial-intelligence-systems/. Under this law, an agency would be prohibited from using an automated decision-making system to discriminate against an individual on the basis of a list of factors such as race, national origin, sex, and age. Id.

The following Gibson Dunn lawyers prepared this client update: H. Mark Lyon, Frances Waldmann, Tony Bedel, Selina Grün, Emily Lamm and Chris Timura.

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work, any member of the firm’s Artificial Intelligence and Automated Systems Group, or the following authors:

H. Mark Lyon – Palo Alto (+1 650-849-5307, [email protected])
Frances A. Waldmann – Los Angeles (+1 213-229-7914,[email protected])

Please also feel free to contact any of the following practice group members:

Artificial Intelligence and Automated Systems Group:
H. Mark Lyon – Chair, Palo Alto (+1 650-849-5307, [email protected])
J. Alan Bannister – New York (+1 212-351-2310, [email protected])
David H. Kennedy – Palo Alto (+1 650-849-5304, [email protected])
Ari Lanin – Los Angeles (+1 310-552-8581, [email protected])
Robson Lee – Singapore (+65 6507 3684, [email protected])
Carrie M. LeRoy – Palo Alto (+1 650-849-5337, [email protected])
Alexander H. Southwell – New York (+1 212-351-3981, [email protected])
Eric D. Vandevelde – Los Angeles (+1 213-229-7186, [email protected])
Michael Walther – Munich (+49 89 189 33 180, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

2019 was another busy year of enforcement for China’s State Administration for Market Regulation (the “SAMR”). On the merger front, the SAMR issued five conditional approval decisions, demonstrating its continued reliance on creative remedies such as hold-separate obligations, commitments to supply products and services on fair, reasonable and non-discriminatory terms and the implementation of antitrust compliance mechanisms. The SAMR’s decision in Novelis’ proposed acquisition of Aleris demonstrated the risks in opting for the simplified review procedure. On the non-merger side, the SAMR imposed heavy penalties on a number of foreign-invested companies for anti-competitive conduct and launched high-profile investigations involving multinational corporations in 2019. Separately, the Supreme People’s Court of China provided valuable insight into the arbitrability of antitrust civil claims and rules on resale price maintenance in two landmark judgments.

2020 is likely to be a very important year for antitrust enforcement China. Since the second half of 2019, the SAMR has been particularly active in publishing new regulations and guidelines, as well as proposing amendments to existing antitrust legislation. In January 2020, the SAMR published proposed amendments to the Anti-Monopoly Law (the “AML”) for comments. The proposed changes include higher fines for merger-related conduct, in response to long-standing concerns that the current fines are too low, as well as a number of other substantive changes.

This client alert highlights the most significant developments from 2019 and what to expect for 2020.

1. Legislative/Regulatory Developments

Since its establishment in March 2018, the SAMR has introduced a number of legislative enactments and guidelines, which not only codify the practices of its predecessors, but also introduce notable changes to the interpretation and enforcement of the AML.

New Antitrust Regulations. The SAMR published three antitrust rules on July 1, 2019: (1) Interim Provisions on Prohibiting Monopoly Agreements;[1] (2) Interim Provisions on Prohibiting Abuse of Market Dominant Position;[2] and (3) Interim Provisions on Restraining Abuse of Administrative Power to Restrict Competition.[3] These regulations entered into force on September 1, 2019 (together, the “Antitrust Regulations”).

The Antitrust Regulations seek to unify the fragmented enforcement scheme that existed in the past by codifying the practice of the SAMR’s predecessors while providing more detailed guidance on certain issues. Furthermore, unlike the regulations issued by the SAMR’s predecessors, the Antitrust Regulations combine both substantive and procedural provisions. Procedurally, the Antitrust Regulations set out the division of responsibility and working mechanisms between the national level and provincial level enforcers and also strengthen the SAMR’s supervision of the provincial level enforcement activity, to ensure consistent standards across the nation. Provincial anti-monopoly enforcement agencies must report to the SAMR within seven business days of initiating an investigation of an alleged AML violation and notify the SAMR before suspending, terminating, or imposing penalties in an investigation.

The Interim Regulation on Prohibition of Monopoly Agreements (the “Monopoly Agreement Regulation”) distinguishes between: (1) typical or “core” monopoly agreements, which are per se illegal; and (2) non-typical monopoly agreements or “non-core” monopoly agreements, which adopt the rule of reason approach and explicitly require the SAMR to analyse whether competition is restricted or excluded as a result. In addition, the Monopoly Agreement Regulation provides clarity on a number of issues. First, it sets out factors for identifying “non-core” monopoly agreements, such as the parties’ market shares and the agreement’s impact on prices and market entry. Second, the Monopoly Agreement Regulation explicitly states that the SAMR will suspend investigations on the basis of commitments for all types of AML violations, save for three types of “hardcore” cartel violations (namely, price fixing, output restrictions and market allocation). Previously, it was unclear whether all types of violations qualified for suspensions or only abuse of dominance cases. Third, the Monopoly Agreement Regulation provides guidance for identifying concerted practices: Article 6 states that the SAMR can determine the existence of a concerted practice after taking into account factors such as uniformity in conduct; exchange of communication or info between the parties; justifications for uniform conduct; and changes in market structure and/or competitive conditions.

Lastly, the Monopoly Agreement Regulation adopts a unified leniency framework for all types of monopoly agreements, including non-hardcore cartels and retail price maintenance. The first applicant with important information may be fully exempted or be given a reduction in fines of 80% or greater. The second applicant may receive a reduction in fines of 30% to 50% and the third application may receive a reduction in fines of 20% to 30%. The main change is in regard to the first applicant: unlike its predecessors, who granted a complete exemption, the SAMR has the discretion to award an exemption or a reduction of fines to the first applicant.

The Interim Provisions on Prohibiting Abuse of Market Dominant Position (the “Abuse of Dominance Provisions”) provides helpful guidance on identifying abuses of a dominant position. In regards to collective dominance, Article 13 of the Abuse of Dominance Provisions states that the SAMR must consider various factors, such as market structure, transparency, homogeneity of the relevant products and uniformity in the conduct of the undertakings. Additional guidance is provided in regards to certain conduct, such as predatory pricing: the SAMR clarified, for the first time, that it will use average variable cost as the appropriate benchmark for predatory pricing. The SAMR will also take into account both paid and complimentary products in predatory pricing cases involving internet companies or new economy ventures. In unfair pricing cases, the SAMR will now also consider as a benchmark the dominant undertaking’s own prices in other geographical markets. Lastly, the Abuse of Dominance Provisions contain a number of defences. For instance, an undertaking may argue that the abusive practice is consistent with the industrial norm; that the abuse practice is required for product safety reasons; or that it is not possible to produce or sell the products without the abusive practice.

Undertakings’ Anti-Monopoly Compliance Guidelines. In November 2019, the SAMR published its draft Undertakings’ Anti-Monopoly Compliance Guidelines (the “Compliance Guidelines”)[4] for public comment. Composed of 6 chapters and 30 articles, the Compliance Guidelines are non-binding and recommend measures for developing and implementing antitrust compliance programs. Many of the Compliance Guidelines are aligned with international best practices for antitrust compliance and similar guidelines from other regulators, such as the DOJ’s July 2019 policy on “Evaluation of Corporation Compliance Programmes in Criminal Antitrust Investigations”.

The Compliance Guidelines adopt a risk-based approach to compliance: companies should identify and assess antitrust risks based on their unique “business conditions, business scale, and industrial features”. The Compliance Guidelines emphasize the importance of instilling a culture of compliance and having senior executives who are openly committed to antitrust compliance. In regards to undertakings that engage in overseas business, the Compliance Guidelines note that these undertakings should understand and abide by the relevant anti-monopoly laws and regulations and obtain advice from antitrust lawyers.

The Compliance Guidelines encourage companies to set up a compliance department that has the resources, independence and capability commensurate with the companies’ compliance risks. This department should understand the relevant provisions of domestic and foreign anti-monopoly laws, formulate internal compliance management measures, requirements and processes and mechanisms for supervising and evaluating the compliance status of the undertaking and its employees. Moreover, this department should liaise with human resources to carry out training and incorporate anti-monopoly compliance into performance evaluation and employee incentives and liaise with the regulator in investigations.

Lastly, the Compliance Guidelines note that companies may voluntarily submit progress reports to the SAMR, setting out the “construction of their anti-monopoly compliance management system” and “the implementation effect”. The benefits of such self-reporting are yet unclear, save for the establishment or maintenance of positive relations with the regulator through open dialogue.

Draft Amendments to the AML. On January 2, 2020, the SAMR released a draft amendment to the AML (the “Draft Amendment”)[5] for public consultation. The deadline for comments was January 31, 2020. The Draft Amendment introduces changes that affect all aspects of AML enforcement, including merger control, non-merger enforcement and procedural rules.

In regards to merger control, the Draft Amendment seeks to increase penalties for breaches of merger-related conduct, including gun-jumping, failure to notify reportable transactions and breach of commitments in conditionally approved cases. The increased fines for gun-jumping are notable. The current version of the AML imposes a maximum fine of RMB 500,000 (~$70,000) for merger-related conduct. The Draft Amendment proposes to change the maximum fine to 10% of the infringing party’s turnover in the previous financial year. Moreover, the Draft Amendment now explicitly defines “control” as the ability to “exert or potentially exert a decisive influence on another undertaking’s production and operation activities or other major decisions”.

In regards to non-merger enforcement, the Draft Amendment also seeks to increase the penalties for anti-competitive agreements that have been entered into but not yet implemented. The Draft proposes a maximum fine of RMB 50 million (~$7 million), which is a 100 fold increase from the previous penalty of RMB 0.5 million (~$70,000). These fines would be applied to even those companies without any turnover in the preceding year and those companies who assisted others in engaging in anti-competitive agreements. Both categories of undertakings are not subject to any penalties under the current version of the AML. In addition, the Draft Amendment widens the scope of enforcement by explicitly including “concerted practices” in its definition of “agreements” and limits the applicability of efficiency claims to agreements that are “indispensable”. Lastly, the Draft Amendment adopts a rule of reason approach for assessing vertical conduct, save for retail price maintenance, which continues to be a per se violation of the AML.

The Draft Amendment also introduces a number of procedural changes to the SAMR’s enforcement of the AML. The Draft Amendment now allows the SAMR to “stop the clock” during a merger investigation (for instance, if the transaction parties supplement their notification with additional materials) and codifies the SAMR’s ability to carry out sub-threshold merger investigations and to amend the turnover notification thresholds from time to time. The Draft Amendment also notes that the SAMR’s investigations into anti-competitive behavior cannot be suspended if they involve hardcore cartel infringements, such as price-fixing, output restriction and market allocation. Finally, under the Draft Amendment, the SAMR may seek assistance from police departments when carrying out an antitrust investigation. This provision is expected to reinforce the SAMR’s ability to conduct dawn raids.

Draft Interim Provisions on the Review of Undertakings Concentrations. On January 7, 2020, the SAMR published the Draft Interim Provisions on the Review of Undertakings Concentrations (the “Draft Merger Review Provisions”)[6] for public consultation. The Draft Merger Review Provisions demonstrate the SAMR’s effort to integrate and streamline rules on its review, investigation and enforcement concerning “concentrations of undertakings,” which govern the merger of undertakings, an undertaking’s acquisition of control over other undertakings by acquiring their equity or assets, and an undertaking’s gaining control and decisive influence over other undertakings by contracts or other means.

In addition to consolidating the existing merger rules and regulations, the Draft Merger Review Provisions provide clarification on key issues such as procedures for regular merger reviews, timeline for investigations, and factors that the SAMR would consider when conducting a review, as well as introduce new measures. Examples of substantive changes include the following:

New Market Share Requirement for Simplified Review Procedure. Article 18 imposes a new market share requirement for transactions to qualify for the simplified review procedure. It provides that where the transaction will result in a joint venture changing from being jointly controlled by two or more undertakings to being controlled by only one of them, and where the controlling undertaking and the joint venture compete in the same market, the transaction will not qualify for the simplified review procedure if the combined market share of the controlling undertaking and the joint venture in the relevant market exceeds 15%.
Additional Divestiture Requirements. Article 47 provides that parties to a proposed concentration should consider naming a specific buyer to purchase assets that they are divesting if the identity of the buyer would have a decisive influence on whether competition in the relevant market could be restored, or if there are circumstances in which the SAMR may consider such proposal appropriate. Article 47 further allows the SAMR to require the undertakings to halt the implementation of the concentration until the divestiture is complete.

The public consultation period for the Draft Merger Review Provisions closed on February 7, 2020, but it is unclear when the Draft Merger Review Provisions will be finalized and come into force.

2. Merger Control

Please refer to the section on China in our Antitrust Merger Enforcement Update and Outlook published on January 29, 2020.[7]

3. Non-Merger Enforcement

In 2019, the SAMR published 19 enforcement decisions, most of which were issued by provincial Administrations for Market Regulation (the “AMRs” or “AMR”). These 19 decisions included 14 penalty decisions, three decisions to suspend investigations and two decisions to terminate investigations.[8] Of the 14 penalty decisions, eight involved horizontal monopolistic agreements, two concerned vertical monopolistic agreements, and the remaining four were abuse of dominance cases.[9]

While these enforcement actions concerned undertakings across industries, including pharmaceuticals, construction, public utilities and automobile, 2019 saw heavy penalties imposed on foreign-invested companies. On April 29, 2019, the SAMR published a decision issued by the Shanghai AMR to impose a fine of RMB 24.38 million (approx. USD 3.51 million) on Eastman (China) Investment Management Co., Ltd. (“Eastman China”), which is a subsidiary of the New York-listed Eastman Chemical Company, for abuse of dominance.[10] The Shanghai AMR found that Eastman China entered into long-term contracts with its customers that included, among other restrictive terms, a minimum purchase obligation and a take-or-pay clause. The fine amounted to 5% of Eastman China’s revenues in 2016.

Two other notable decisions involving heavy fines imposed on foreign-invested companies concerned the automobile industry. In 2019, a Chongqing-based automotive manufacturing company, which is a joint venture between a Chinese company and one of the world’s largest automaker that is listed on the New York Stock Exchange, was fined a total of RMB 162.8 million (approx. USD 23.47 million) by the SAMR . The SAMR found that the joint venture company sought to impose minimum resale prices on its dealers in Chongqing since 2013 by, among other conduct, implementing price lists. The fine amounted to 4% of the joint venture company’s revenues in 2018.

Towards the end of the year, the Jiangsu AMR issued a decision to impose a fine of RMB 87.6 million (approx. USD 12.63 million) on the Chinese subsidiary of a foreign car manufacturing company for engaging in price-fixing conduct. The fine accounted for 2% of the Chinese company’s revenues in 2016. The Jiangsu AMR found that between 2015 and 2018, the Chinese company deprived local car dealers of pricing autonomy by setting minimum resale prices for its cars in Jiangsu Province. The regulators also found that the Chinese company imposed various price-control measures to ensure that the car dealers would abide by the minimum resale prices, including cutting supplies to dealers who sold the cars at lower prices.

In addition to heavy penalties imposed on foreign-invested companies, the SAMR has launched high-profile investigations involving multinational corporations in 2019. These investigations will likely continue through 2020.

4. Civil Litigation

In 2019, The Supreme People’s Court of China (the “SPC”) handed down landmark judgments in cases concerning the arbitrability of antitrust civil claims and rules on resale price maintenance.

Arbitrability of Civil Claims. In China, the arbitrability of private antitrust actions has long been a highly debated topic among practitioners and academics. The controversy largely stems from the fact that both the Arbitration Law and the AML are silent on the matter, and that Chinese courts have taken inconsistent approaches over the years.

The inconsistency has seemingly been resolved when the SPC handed down the judgment in Shell (China) Limited v Hohhot Huili Material Co., Ltd. on August 21, 2019. Hohhot Huili Material Co., Ltd. (“Huili”), one of Shell (China) Limited’s (“Shell China”) distributors in China, alleged that Shell China organized several of its other distributors to enter into a horizontal monopolistic agreement and collude in the bidding process. Shell China argued that Chinese courts did not have jurisdiction to adjudicate the matter because Shell China and Huili agreed in their distribution agreement to settle disputes by arbitration.

The SPC rejected Shell China’s arguments and held that a contractual arbitration clause cannot exclude the Chinese courts’ jurisdiction in adjudicating antitrust civil disputes. The SPC first observed that according to the AML, the question of whether certain conducts are monopolistic should be determined and handled through either administrative enforcement or civil litigation. More importantly, the SPC pointed out that the AML makes no reference to resolving anti-monopoly issues by arbitration.

The SPC then examined the scope of arbitrable disputes under the Arbitration Law. The SPC noted that Article 2 of the Arbitration Law provides that “contractual disputes and disputes arising from property rights may be put to arbitration.” According to the SPC, it follows that Chinese courts would have jurisdiction over disputes that do not arise from contractual or other property rights, or ones that involve elements beyond the scope of arbitrable disputes.

Turning to the facts of the case, the SPC noted that the case was an anti-monopoly dispute rather than a contractual dispute. Although Shell China and Huili agreed to resolve disputes by arbitration, the SPC held that the AML falls within the sphere of public law, as the purpose of the legislation is to, among other aims, prevent and halt monopoly behaviour, maintain fair competition in the market and protect consumer interest and public interest in society. As a result of this classification, the determination of whether certain conducts are monopolistic is a matter that goes beyond the rights and obligations of contractual parties. In other words, anti-monopolistic disputes are not within the scope of arbitrable disputes as defined in the Arbitration Law. Therefore, the SPC held that Shell China could not rely on the arbitration clause in the distribution contract to exclude the court’s jurisdiction to adjudicate the dispute.

New Analysis Framework on Resale Price Maintenance. On June 24, 2019, the SPC published its December 2018 ruling in Hainan Provincial Price Bureau v. Hainan Yutai Scientific Feed Company, which is a judicial review concerning resale price maintenance (“RPM”). This landmark decision addressed the longstanding conflicting approaches on RPM analysis taken by the Chinese antimonopoly authorities and the Chinese courts: while antimonopoly authorities treated RPM agreements as illegal per se, Chinese courts adopted a rule of reason approach in cases such as Dongguan Hengli Guochang Electronics Store v. Dongguan Yushi Xinqing Geli Trading Co., Ltd. and Dongguan Heshi Electronics Co., Ltd. (more commonly known as the Gree case).

In February 2017, the Hainan Provincial Price Bureau (the “Price Bureau”) imposed a fine of RMB 200,000 (approx. USD 28,832) on Hainan Yutai Scientific Feed Company (“Yutai”) for concluding agreements with its distributors in 2014 and 2015 that required the distributors to follow “guiding prices” set by Yutai in their resale to third parties. Yutai filed for judicial review of the decision and won on first instance. The Price Bureau successfully appealed the first instance decision, and Yutai further appealed to the SPC in July 2018.

The SPC rejected Yutai’s appeal and clarified a number of key issues on the enforcement of the rules governing RPM under Article 14 of the AML. The court first provided an overview of Article 13 of the AML, which introduces the concept of a “monopolistic agreement,” defined as “agreements, decisions or concerted actions which eliminate or restrict competition” in the AML. The court explained that certain agreements, decisions or concerted actions, once formed, will necessarily lead to the elimination or restriction of competition, and thus are illegal per se. According to the SPC, price-fixing, output restriction and market-sharing are commonly accepted as unreasonable restrictive behaviours that are harmful to competition. As a result, antimonopoly authorities are entitled to presume such behaviors as unlawful once there is sufficient evidence to prove the existence of such behaviors, without having to engage in a comprehensive analysis on whether they eliminate or restrict competition.

Then SPC then examined Article 14 of the AML, which prohibits three types of agreements between an undertaking and its trade counterparts: (1) agreements that fix the resale price; (2) agreements that restrict the lowest resale price; and (3) other monopolistic agreements identified by the antimonopoly authorities. Noting that the first and second type of agreements are typical vertical monopolistic agreements, which fall within the scope of “monopolistic agreements” as defined in Article 13 of the AML, the SPC held that the Chinese antimonopoly authorities need not prove anticompetitive effects in cases where they have established the existence of such agreements. In other words, there is a presumption that RPM agreements eliminate or restrict competition and are thus illegal. The SPC added that this presumption is rebuttable by the undertaking adducing sufficient evidence to either prove that the RPM agreement does not eliminate or restrict competition, or demonstrate that the agreement falls within one of the exemptions set out in Article 15 of the AML.[11]

In explaining its rationale for placing a presumption of unlawfulness on RPM in favor of the Chinese antimonopoly authorities, the SPC acknowledged that RPM can have both procompetitive and anticompetitive effects. However, given the relatively immature market conditions in China and the relatively weak self-correction tendencies of the Chinese marketplace, it is essential that the antimonopoly authorities are empowered to focus on regulating and penalizing RPM agreements. The SPC determined that it would not be appropriate to require antimonopoly authorities to conduct a comprehensive investigation and engage in complex economic assessment in every RPM case, which would only substantially raise the costs and lower the efficiency of antimonopoly enforcement.

On the facts of the case, the SPC found that Yutai failed to provide sufficient evidence to demonstrate that the RPM agreements that it concluded with its distributors did not restrict or eliminate competition in the relevant market. In particular, the SPC rejected the first instance court’s finding that the agreements had no anticompetitive effects due to Yutai’s relatively low sales volume and market share. The SPC criticized the first instance court for failing to engage in any comprehensive analysis or examination of evidence to support its conclusion.

Finally, the SPC drew a distinction between administrative enforcement cases, such as the present case, and civil disputes. The SPC held that plaintiffs in civil litigation will not be entitled to the presumption that RPM agreements are unlawful. Instead, in order to recover monetary damages, plaintiffs must prove that they actually suffered harm, which necessarily involves a demonstration of the RPM agreement’s anticompetitive effects.

Shortly after the publication of this SPC decision, the SAMR issued the Interim Regulation on Prohibition of Monopoly Agreements (the “Monopoly Agreement Regulation”) on July 1, 2019. The Monopoly Agreement Regulation provide that agreements specifically enumerated in Article 14 of the AML, namely agreements that fix the resale prices or restrict the lowest resale prices, are deemed illegal per se. Given that the SPC decision has now been enacted into law and that there is little guidance as to the types of evidence that would be considered sufficient to rebut the presumption that RPM agreements are unlawful, undertakings are advised to be very cautious before imposing any pricing requirements on their distributors or retailers.

______________________

[1] SAMR, “Interim Provisions on Prohibiting Monopoly Agreements” (禁止垄断协议暂行规定) (released on July 1, 2019), available at http://gkml.samr.gov.cn/nsjg/fgs/201907/t20190701_303056.html.

[2] SAMR, “Interim Provisions on Prohibiting Abuse of Market Dominant Position” (禁止滥用市场支配地位行为暂行规定) (released on July 1, 2019), available at http://gkml.samr.gov.cn/nsjg/fgs/201907/t20190701_303057.html.

[3] SAMR, “Interim Provisions on Restraining Abuse of Administrative Power to Restrict Competition” (制止滥用行政权力排除、限制竞争行为暂行规定) (released on June 26, 2019), available at http://www.gov.cn/gongbao/content/2019/content_5433728.htm.

[4] SAMR, “Announcement of the State Administration for Market Regulation to Seek Public Comment on the Undertakings’ Anti-Monopoly Compliance Guidelines (Draft for Public Comment)” (市场监管总局关于就《经营者反垄断合规指南（公开征求意见稿）》公开征求意见的公告) (released on November 28, 2019), available at http://www.samr.gov.cn/hd/zjdc/201911/t20191128_308890.html.

[5] SAMR, “Announcement of the State Administration for Market Regulation to Seek Public Comment on the Revised Draft of the Anti-Monopoly Law (Draft for Public Comment)” (市场监管总局就《<反垄断法>修订草案（公开征求意见稿）》公开征求意见的公告) (released on January 2, 2020), available at http://www.samr.gov.cn/hd/zjdc/202001/t20200102_310120.html.

[6] SAMR, “Announcement of the State Administration for Market Regulation to Seek Public Comment on the Interim Provisions on the Review of Undertakings Concentrations (Draft for Public Comments)” (市场监管总局关于《经营者集中审查暂行规定（征求意见稿）》公开征求意见的公告) (released on January 7, 2020), available at http://www.samr.gov.cn/fldj/tzgg/zqyj/202001/t20200107_310322.html.

[7] Gibson, Dunn & Crutcher, “Antitrust Merger Enforcement Update and Outlook” (released on January 29, 2020), available at https://www.gibsondunn.com/antitrust-merger-enforcement-update-and-outlook-january-2020/.

[8] SAMR, “Administrative Penalty Cases” (行政处罚案件), available at http://www.samr.gov.cn/fldj/tzgg/xzcf/.

[9] SAMR, “Administrative Penalty Cases” (行政处罚案件), available at http://www.samr.gov.cn/fldj/tzgg/xzcf/.

[10] SAMR, “The State Administration for Market Regulation Published Administrative Penalty Decision Against Eastman Corporation for Abuse of Market Dominance” (市场监管总局发布伊士曼公司滥用市场支配地位案行政处罚决定书) (released on April 29, 2019), available at http://www.samr.gov.cn/fldj/tzgg/xzcf/201904/t20190429_293241.html.

[11] Article 15 of the AML provides for exemptions where the objective of the monopolistic agreement is to, among other societal benefits, improve technology, enhance the competitiveness of small and mid-sized enterprises or achieve public interests.

The following Gibson Dunn lawyers assisted in the preparation of this client update: Sébastien Evrard, Emily Seo and Bonnie Tong.

Gibson Dunn lawyers are available to assist in addressing any questions you may have regarding these issues. Please contact the Gibson Dunn lawyer with whom you usually work, any member of the firm’s Antitrust and Competition Practice Group, or the following lawyers in the firm’s Hong Kong office:

Sébastien Evrard (+852 2214 3798, [email protected])
Kelly Austin (+852 2214 3788, [email protected])
Emily Seo (+852 2214 3725, [email protected])
Bonnie Tong (+852 2214 3762, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

Los Angeles partner Abbey Hudson and associates Dione Garlick and Caroline Monroy are the authors of “Calif. Low Carbon Fuel Standard Price Cap Is A Trade-Off,” [PDF] published by Law360 on February 7, 2020.

Orange County partner Blaine Evanson and Washington, D.C. associate Jeremy Christiansen are the authors of “Rimini V. Oracle’s Ripple Effect On Textualism, Expenses,” [PDF] published by Law360 on February 6, 2020.

On January 30, 2020, the five regulatory agencies (Agencies) responsible for implementing the Dodd-Frank Act’s Volcker Rule issued, in some cases with dissent, a proposed rule (Funds Proposal) that would make substantial revisions to the “covered funds” provisions of their Volcker regulations. The proposal would loosen a number of the restrictions imposed by the original 2013 regulation (Original Rule) and add new exclusions from the Rule’s prohibitions.

Of particular significance, the Funds Proposal would:

Exclude “credit funds” from the Volcker Rule, which would allow banking entities to invest their own money in such funds without limitation
Exclude “venture capital funds” from the Volcker Rule, which would allow banking entities to invest their own money in such funds without limitation
Eliminate restrictions on banking entities’ directly investing their own money in parallel with investments by covered funds
Permit certain exceptions to the so-called “Super 23A” provisions, which limit transactions between banking entities and covered funds that they advise or sponsor
Exclude so-called “foreign excluded funds” from the Volcker Rule’s prohibitions on proprietary trading and fund investments
Permit exempt loan securitization vehicles to hold up to 5% of their assets in previously impermissible investments

In each of the above cases, the Agencies made considerable use of their exemptive authority in clear contrast to their approach in the Original Rule.

Comments on the Funds Proposal are due on April 1, 2020.

I. Proposed New Exclusion for Credit Funds

The Funds Proposal would provide a new exclusion from the definition of Volcker “covered fund” for “credit funds.” The effect of the exclusion would be that banking entities could invest in up to 100% of the interests of such funds. Although the Agencies had rejected such an exclusion in the Original Rule, they stated that time had shown that credit funds had had difficulty fitting within any other exclusions, and that allowing the exclusion would be consistent with Congress’s intent that the Volcker Rule not restrict banks’ ability to sell loans.

The Funds Proposal defines a “credit fund” as an issuer whose assets solely consist of: (1) loans; (2) debt instruments; (3) related rights, and other assets that are related to or incidental to acquiring, holding, servicing, or selling loans or debt instruments – including warrants and other “equity kickers;” and (4) certain interest rate and foreign exchange derivatives. Permissible related rights and derivatives would be similar to those allowed in the regulations’ loan securitization exclusion. With respect to warrants and equity kickers, the Agencies stated that they were considering imposing a quantitative limit on the amount of equity a credit fund could hold and sought comment on the issue.

Although the exclusion would mean that a banking entity could invest in up to 100% of the interests of a credit fund, regardless of whether the banking entity or a third party was the fund sponsor, the Funds Proposal does place certain limitations on a credit fund:

A credit fund is not permitted to engage in proprietary trading
If a banking entity sponsors or serves as an investment adviser to a credit fund, it must provide the type of disclosures to investors that it would provide for a covered fund
A banking entity cannot guarantee the performance of a credit fund
Super 23A (as modified) would apply to transactions between a sponsoring or advising banking entity and a credit fund
The prohibition on material conflicts of interest and high-risk trading strategies would apply to a credit fund
A credit fund may hold only those assets that a bank can hold directly, and thus may hold equity only as a “kicker” to a loan, not equity generally
A credit fund is not permitted to issue asset-backed securities

II. Proposed New Exclusion for Venture Capital Funds

The Funds Proposal also contains a new exclusion for qualifying venture capital funds. As with credit funds, the Original Rule declined to provide such an exclusion. The Funds Proposal quotes at length from the legislative history of the Dodd-Frank Act, in particular, floor statements from members of Congress, as justification for the new exclusion.

A “qualifying venture capital fund” is an issuer that meets the requirements set forth in the definition of “venture capital fund” contained in Rule 203(l)-1 under the Investment Advisers Act and meets the additional conditions for credit funds listed above, other than the prohibition on issuing asset-backed securities and the limitation of fund assets to those assets that a bank can hold directly.

Under Rule 203(l)-1, a “venture capital fund” is limited in the types of companies in which it can invest (private portfolio companies), the types of assets it may hold (principally equity of qualifying portfolio companies and short-term assets) and the amount of leverage that it may incur. The Agencies stated that they were further considering imposing a limitation on the amount of annual revenues that could be generated by portfolio companies in which a qualifying venture capital fund invests (for example, $50 million in annual revenues) and sought comment on that issue.

III. Eliminating Restrictions on Direct Investments by Banking Entities and Their Officers and Directors

In a substantial change relating to bank investments, the Funds Proposal would permit banking entities to make parallel direct investments in portfolio companies alongside investments by sponsored covered funds, without counting the direct investments towards the 3% per-fund and 3% of Tier 1 capital investment limits. The preamble to the Funds Proposal states further that banking entities could market sponsored funds by referring to a direct co-investment strategy.

In addition, the loosening of restrictions on parallel investments would apply to director and employee investments. Such investments would not be attributed to the 3% per-fund and 3% of Tier 1 capital limits, and could be made by bank directors and employees that provided no services to the fund.

IV. Revisions to Super 23A Provisions

One of the more limiting aspects of the Volcker Rule is its so-called “Super 23A” provision, which places outright prohibitions on certain transactions between banking entities and covered funds that they sponsor or advise – extensions of credit, guarantees issued on behalf of the fund, and purchases of assets or securities from the fund. One of the reasons the provision is called “Super 23A” is that the Volcker statute did not include any language stating that the exemptions in Section 23A of the Federal Reserve Act, on which Super 23A is based, should apply. In the Original Rule, the Agencies declined to import the Federal Reserve Act Section 23A exemptions by administrative action.

The Funds Proposal changes course on this issue as well. It includes, moreover, not simply the historical Section 23A exemptions, but an additional exemption as well. The principal Section 23A exemptions are for: (i) extensions of credit secured by government securities or a segregated, earmarked deposit account; (ii) purchases of certain assets having a publicly available and readily identifiable market quotation and purchased at that quotation; (iii) purchases of certain marketable securities; (iv) purchasing certain municipal securities; and (v) intraday extensions of credit. The additional proposed exemption is for an extension of credit to, or purchase of asset from, a covered fund, as long as:

The transaction is in the ordinary course of business in connection with payment transactions; settlement services; or futures, derivatives, and securities clearing
Each extension of credit is repaid, sold or terminated by the end of five business days, and
The banking entity making each extension of credit has established and maintains policies and procedures reasonably designed to manage the credit exposure arising from the extension of credit in a safe and sound manner and ensure that it is on market terms, and has no reason to believe that the covered fund will have difficulty repaying the extension of credit in accordance with its terms.

V. “Foreign Covered Funds” and the “Banking Entity” Problem

The Original Rule created a substantial issue for non-U.S. banking organizations. The statutory prohibitions on proprietary trading and investing in hedge funds and private equity funds apply broadly to every subsidiary in a bank holding company structure, because those subsidiaries are “banking entities” subject to the prohibitions. The Original Rule exempted “covered funds” from the banking entity definition, since the statute permitted banking entities to sponsor hedge funds and funds of funds, which by definition engage in proprietary trading and fund investing.

The “covered fund” definition in the Original Rule, however, did not cover many funds sponsored by non-U.S. banks, i.e, those that had no U.S. investors. Because they were sponsored by non-U.S. banks, these “foreign excluded funds” were controlled companies and thus “banking entities” that were technically prohibited from proprietary trading and investing in private funds. The Volcker Agencies did not attempt to revise their regulations when this became apparent; rather, once the Volcker Rule conformance period finally ended in July 2017, the federal banking agencies issued annual orders that they would not take any action with respect to foreign excluded funds – essentially deferring the issue.

In what, if finalized, would be a substantial victory for non-U.S. banks, the Covered Funds Proposal would exempt qualifying “foreign excluded funds” from the prohibitions on proprietary trading and sponsoring and investing in hedge funds and private equity funds. To qualify for the exemption, a fund must meet these requirements:

Be organized and established outside the United States, with all ownership interests offered and sold solely outside the United States
Would be a covered fund if the entity were organized or established in the U.S., or is or holds itself out as being, an entity or arrangement that raises money from investors primarily for the purpose of investing in financial instruments for resale or other disposition or otherwise trading in financial instruments
Would not otherwise be a banking entity except by virtue of a foreign banking entity’s acquisition or retention of an ownership interest in, or sponsorship of, the entity
Be established and operated as part of a bona fide asset management business
Not be operated in a manner that enables evasion of the requirements of the Volcker Rule.

If finalized as proposed, this exemption will significantly reduce compliance costs for non-U.S. banks.

VI. Broadening the Types of Assets That Can Be Held by Loan Securitizations

The Volcker statute provided that “[n]othing [herein] shall be construed to limit or restrict the ability of a banking entity . . . to sell or securitize loans in a manner otherwise permitted by law.”[1] Based on this legal authority, the Original Rule contained an exemption from the definition of “covered fund” for loan securitization vehicles; those vehicles, however, were significantly limited in the types of assets they could hold. For example, as a general matter, debt securities were impermissible assets.

The Funds Proposal reverses this position. Reasoning that authorizing loan securitizations to hold small amounts of non-loan assets could “permit loan securitizations to respond to market demand and reduce compliance costs” without “significantly increasing risk to banking entities and the financial system” and also “increase a banking entity’s capacity to provide financing and lending,” the Funds Proposal would allow a loan securitization vehicle to hold up to five percent of assets in otherwise impermissible assets.

VII. Other Proposed Changes

Family Wealth Management Vehicles. The Funds Proposal would include a new exclusion from the definition of covered fund for “family wealth management vehicles.” Such vehicles would not hold themselves out as raising money from investors generally; they could be either trusts, where all the grantors were family customers, or non-trust vehicles, where a majority of the voting interests were owned by family customers, and the entity was owned only by family customers and up to 3 closely related persons of those family customers. A “family customer” would mean “a family client, as defined in Rule 202(a)(11)(G)-1(d)(4) of the Advisers Act; or . . . any natural person who is a father-in-law, mother-in-law, brother-in-law, sister-in-law; son-in-law or daughter-in-law of a family client, spouse or spousal equivalent of any of the foregoing.”

In addition, Super 23A, as modified, would not apply to transactions between a banking entity and a sponsored or advised “family wealth management vehicle,” but the Agencies would impose the limitation that a banking entity could not purchase a low-quality asset from such vehicles. Such vehicles would also be subject to the following requirements and limitations:

Banking entity transactions with the vehicles would be required to be on market terms
Banking entity ownership would be limited to 0.5% and only when necessary for establishing corporate separateness or to address bankruptcy/insolvency concerns
Banking entities would be required to provide the same type of disclosures to vehicle investors as they would covered funds
Banking entities could not guarantee the obligations or performance of the vehicles
The prohibitions against material conflicts of interest and high-risk trading strategies would apply

Customer Facilitation Vehicles. The last new exclusion from the definition of covered fund would be for “customer facilitation vehicles.” This exclusion would cover any issuer that is formed “by or at the request of” a customer of a banking entity for the purpose of providing the customer or its affiliates with exposure to a transaction, investment strategy or other service provided by the banking entity. A banking entity could market its services through the use of customer facilitation vehicles and discuss with customers prior to formation of the vehicle the potential benefits of using such a vehicle. Such vehicles would be subject to the same conditions as family wealth management vehicles.

Definition of Ownership Interest. The Funds Proposal would eliminate an inconsistency in the manner in which the 3% limits and the regulations’ capital deduction are calculated, by requiring banking entities to include employee/director payments in connection with a “restricted profits interest” (carried interest) only when the banking entity had financed such payments. This would result in a uniform approach to insider investments.

The Agencies also included clarifications to the definition of “ownership interest” so that a debt relationship with a covered fund would typically not constitute an ownership interest.

Foreign Public Funds. The Funds Proposal would loosen some of the conditions for the exclusion from the covered fund definition for “foreign public funds,” namely, that the fund be authorized to be offered and sold to retail investors in the fund’s home jurisdiction, and that the fund be sold “predominantly” through one or more public offerings. In addition, it would eliminate the limitation on selling ownership interests in the fund to employees, other than senior executive officers and directors, of the sponsoring banking entity, the fund, or their affiliates. The Agencies stated that these changes would align the treatment of foreign public funds more closely with exempt U.S. registered investment companies.

Public Welfare Funds. The Agencies sought comment on whether the exclusion from the definition of covered fund should cover “all permissible public welfare investments,” under any banking agency’s regulation, as well as comment on how the exclusion should apply to Rural Business Investment Companies and Qualified Opportunity Funds.

Small Business Investment Companies. The Funds Proposal would permit a small business investment company that had surrendered its license to continue to benefit from the SBIC exclusion from the definition of covered fund as long as it did not make any further portfolio investments.

Conclusion

The Funds Proposal would result in a material revision to the Volcker Rule. It shows significant flexibility on the part of the Agency staffs in revisting their prior positions in the Original Rule, flexibility that is not generally associated with regulators. Indeed, there is some irony that the Funds Proposal was issued on the same day that the Federal Reserve finalized its revised framework for “control” under the Bank Holding Company Act, which made rather restrained amendments to prior practice. The Volcker Agencies seem to have determined, after a little over six years’ experience, that bank fund activities may be broadened, in some cases, substantially, without creating issues of undue risk taking.

______________________

[1] 12 U.S.C. § 1851(g)(2).

Gibson Dunn’s lawyers are available to assist in addressing any questions you may have regarding these developments. Please contact the Gibson Dunn lawyer with whom you usually work in the firm’s Financial Institutions or Investment Funds practice groups, or the following:

Financial Institutions Group:
Arthur S. Long – New York (+1 212-351-2426, [email protected])
Matthew L. Biben – New York (+1 212-351-6300, [email protected])
Michael D. Bopp – Washington, D.C. (+1 202-955-8256, [email protected])
Stephanie Brooker – Washington, D.C. (+1 202-887-3502, [email protected])
M. Kendall Day – Washington, D.C. (+1 202-955-8220, [email protected])
Michelle M. Kirschner – London (+44 20 7071 4212, [email protected])
Jeffrey L. Steiner – Washington, D.C. (+1 202-887-3632, [email protected])
James O. Springer – New York (+1 202-887-3516, [email protected])

Investment Funds Group:
Chézard F. Ameer – Dubai and London (+971 (0)4 318 4614, [email protected])
Albert S. Cho – Hong Kong (+852 2214 3811, [email protected])
John Fadely – Hong Kong (+852 2214 3810, [email protected])
Jennifer Bellah Maguire – Los Angeles (+1 213-229-7986, [email protected])
Y. Shukie Grossman – New York (+1 212-351-2369, [email protected])
Edward D. Sopher – New York (+1 212-351-3918, [email protected])
C. William Thomas, Jr. – Washington, D.C. (+1 202-887-3735, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

New York associate Alex Marcellesi is the author of “Guaranteed Payments and Interest: Why Treasury Is Overreaching,” [PDF] published by Tax Notes Federal on December 9, 2019.

As a companion to our recent alert on considerations for preparing your 2019 Form 10-K (available here), below we offer our observations on new developments and recommended practices to consider in preparing the 2020 Proxy Statement. In particular, there are a number of important substantive and technical considerations that public companies should keep in mind in light of changes to U.S. Securities and Exchange Commission (“SEC”) rules, developments relating to institutional investor and proxy advisory firm policies, and the continuing evolution of corporate governance best practices.

Address New Hedging Disclosures and Consider in Advance the Need to Revise Related Policies

Many companies already provide proxy statement disclosure about their hedging policies, particularly as applicable to directors and executive officers, but these disclosures may need to be expanded in light of the new SEC rule requiring hedging policy disclosures. Companies also should review their policies and determine whether any changes are appropriate.

In December 2018, the SEC approved final rules implementing Section 955 of the Dodd-Frank Wall Street Reform and Consumer Protection Act of 2010, which requires companies to disclose their practices or policies with respect to hedging transactions by directors, officers and other employees. The new disclosures, found in Item 407(i) of Regulation S-K, are required for proxy statements that relate to the election of directors during fiscal years beginning on or after July 1, 2019; however, there is a one-year delay for smaller reporting companies and emerging growth companies.

Pursuant to the final rules, companies must provide a summary of their practices or policies (whether written or not) or disclose them in full. The disclosure must state who is covered by the hedging practices or policies and which categories of hedging transactions are specifically permitted or specifically disallowed (including the company’s determinations about what activities are covered by the practices or policies). If a company does not maintain policies or practices on hedging, it must disclose this fact or affirmatively state that hedging transactions generally are permitted. According to a review of the first 40 proxy statements that included the newly required disclosures, every company disclosed that it had a hedging policy in place and 62% had hedging policies that cover both directors and all employees.[1] The final rules largely sidestep difficult definitional questions by providing that companies may use their own policies’ and practices’ terms and the categories of persons covered. Companies need not identify every type of transaction that is permissible if not specifically addressed in their policies or practices and need not address categories of transactions or persons that are not covered.[2]

Placement of the disclosure also should be considered. Item 407(i) does not specify where the disclosure should appear in the proxy statement, so companies have flexibility in this regard. At many companies, the Compensation Discussion and Analysis (“CD&A”) already addresses hedging policies as they apply to named executive officers (“NEOs”). In light of this, in the release adopting Item 407(i), the SEC noted that companies have three options:[3]

include the new hedging disclosure outside the CD&A and provide separate disclosure for NEOs in the CD&A;
include the new hedging disclosure outside the CD&A and incorporate it into the CD&A with a cross-reference; or
include the new hedging disclosure in the CD&A.

According to the review (mentioned above) of the first 40 proxy statements filed under the new rules, 60% of companies included their hedging disclosures only in the CD&A, 15% included these disclosures only somewhere other than the CD&A, and 25% included the disclosure in both the CD&A and elsewhere in the proxy statement. Some companies may not wish to include the new hedging disclosure in the CD&A, as it would result in CD&A disclosure about compensation policies for individuals other than NEOs and make the disclosure subject to the say-on-pay vote. As an alternative, companies could include the new disclosure in a proxy section discussing general corporate governance best practices or next to the pay ratio disclosure.

Consider Bolstering the Description of Audit Committee Oversight

Companies should consider bolstering the description of the audit committee’s work in light of the SEC’s December 2019 statement on audit committees. The statement, which was released by SEC Chairman Jay Clayton, Chief Accountant Sagar Teotia, and the Director of the Division of Corporation Finance, William Hinman, addresses the role of the audit committee in financial reporting and highlights key reminders regarding audit committee oversight responsibilities.[4]

In recent years, many companies have voluntarily expanded their proxy statement disclosures about the role and responsibilities of the audit committee. For example, there has been an uptick in disclosures about the committee’s role in selecting the lead audit partner and factors the committee considers in evaluating the qualifications and performance of the outside auditor. Although the SEC’s December 2019 statement covers a range of topics (as discussed in more detail in our blog[5]), a theme that runs through the observations is an emphasis on active engagement by the audit committee. Companies can use the proxy statement to highlight key practices the audit committee uses to engage proactively on these topics. This could include disclosure about how the committee:

sets the “tone at the top” and emphasizes the importance of an environment that supports integrity in the financial reporting process;
oversees processes for monitoring auditor independence;
oversees implementation of new accounting standards;
oversees and participates in the resolution of internal control issues, where identified;
communicates with the outside auditor on matters related to the conduct of the audit and on critical audit matters expected to be described in the auditor’s report; and
reviews and understands non-GAAP measures, and related company policies and disclosure controls.

Prepare for and Address New Institutional Investor and Proxy Advisory Firm Policies

Companies should review key investor and proxy advisory firm policies and evaluate their disclosures in light of those policies. Policy changes can occur even once proxy season is underway, which happened in 2019 with the “overboarding” policies of some institutional investors. Accordingly, it is important for companies to check regularly for policy updates in case it is necessary to respond by filing additional soliciting materials. As of the date of this alert, Vanguard has issued its 2020 proxy voting policies (with minimal changes from 2019), but 2020 proxy voting policies for BlackRock and State Street have not been published (although each has announced their priorities in 2020 with respect to ESG issues, as discussed below).[6]

Changes in overboarding policies had a discernible impact on support levels for directors in 2019,[7] so companies should pay particular attention to this area in 2020. The table below summarizes the current overboarding thresholds of certain major institutional investors and proxy advisory firms, according to their respective voting policies, but companies should familiarize themselves with the thresholds of all their major stockholders. If a director exceeds the relevant limits, steps can be taken (such as stepping off a board or publicly committing to do so) to avoid an adverse vote or voting recommendation. Notably, as part of the updates to its 2020 voting policies, Vanguard appears to have introduced some flexibility and may consider supporting directors who exceed its limits “because of company-specific facts and circumstances that indicate the director will indeed have sufficient capacity to fulfill his/her responsibilities.”

Institution	Maximum Public Company Boards Permitted
BlackRock	2 public company boards for directors who are CEOs of public companies (only 1 outside board other than where they are CEO) 4 public company boards for all other directors
Vanguard	2 public company boards for directors who are NEOs of public companies (only 1 outside board other than where they are an NEO) 4 public company boards for all other directors
State Street	3 public company boards for directors who are CEOs of public companies 6 public company boards for all other directors
ISS	3 public company boards for directors who are CEOs of public companies (2 outside boards other than where they are CEO) 5 public company boards for all other directors
Glass Lewis	2 public company boards for directors who are executives of public companies 5 public company boards for all other directors

In light of continued investor focus on board diversity, consideration should also be given to adopting and disclosing a “Rooney Rule” (named for an NFL policy requiring every team with a head coaching vacancy to interview at least one or more diverse candidates). A robust version of the Rooney Rule would state that as part of the search process for each new director, the nominating/corporate governance committee includes women and minorities in the pool of candidates and instructs search firms to do so. Boards could go further and also commit to interviewing at least one woman and one minority director candidate.

Companies may also wish to provide additional specificity about director attendance at board and committee meetings. Under a 2020 update to its voting policies, Glass Lewis will generally recommend votes “against” the chair of the nominating/corporate governance committee “when directors’ records for board and committee meeting attendance are not disclosed,” or when a company discloses that one or more directors attended less than 75% of board and committee meetings but it is not possible to determine who those directors are. SEC rules already require companies to name any directors who fall below this 75% threshold, but the rules do not require any disclosures about meeting attendance for directors who meet this threshold. In explaining the reason for this new policy, Glass Lewis states that “attendance at board and committee meetings is one of the most basic ways for directors to fulfill their responsibilities to shareholders and that disclosure of attendance records is a critical element in evaluating the performance of directors more generally.” It is not clear what level of specificity Glass Lewis expects to see under this policy. Some companies already include a general statement about directors’ aggregate attendance at meetings in the proxy statement—for example, a statement to the effect that each director attended at least x% of board and applicable committee meetings held during the fiscal year. At a minimum, companies should consider including this type of affirmative statement, in addition to evaluating whether more detailed disclosure would be appropriate.

Finally, companies should be aware of changes at Glass Lewis that will impact pay-for-performance evaluations and say-on-pay voting recommendations in 2020. The changes are the result of Glass Lewis’s transition, effective January 1, 2020, to CGLytics as its exclusive global partner for compensation data and analytics. According to Glass Lewis, because of the transition, peer companies in 2020 will be “significantly different” from those used under the prior methodology. This, in turn, will result in changes to both pay-for-performance grades and voting recommendations at individual companies. At the market-wide level, there will be no changes to the distribution of grades or against voting recommendations because grades are dispersed evenly under the pay-for-performance methodology.

Review Prior Proxy Advisory Firm Reports and Shareholder Engagement Takeaways

Companies should revisit their 2019 ISS and Glass Lewis reports, as well as takeaways from their off-season engagements with investors, in order to identify any matters that should be addressed or given further emphasis in the proxy statement. That way, if ISS or Glass Lewis included cautionary language in a company’s report, or criticized aspects of the company’s corporate governance or executive compensation practices, or if a company received feedback from major stockholders about a particular practice, the company can address this issue appropriately. The company can then describe steps taken or changes made, and any mitigating factors, in the proxy statement.

As an example, beginning in 2020, ISS will begin recommending votes “against” board members who are responsible for approving director compensation where there is a “pattern,” for two or more years, of awarding “excessive” compensation without disclosure of a “compelling rationale” or other mitigating factors. ISS delayed the application of this policy for a year beyond its planned 2019 effectiveness, but in 2019 included cautionary language in its reports for companies where director compensation raised concerns under this policy. Companies that received cautionary language under this policy in 2019 will need to expand their director compensation disclosures to add the “compelling rationale” ISS expects to see. This rationale should provide a reasoned explanation for why the board’s director compensation levels are appropriate, including addressing topics such as the philosophy behind the director compensation program, how the program aligns directors’ interests with those of shareholders, how compensation is linked to directors’ responsibilities and expertise, the mix of compensation, how the board and responsible committee make decisions about director compensation and how director pay levels compare to peers. As director compensation remains a subject of shareholder litigation, there is value in addressing these considerations in the proxy statement, whether or not ISS identifies a company as having “excessive” director compensation.

Consider Adding or Expanding Discussion of “Hot Topics” Like Human Capital Management and Other ESG Topics

Companies should consider adding or expanding discussion in the proxy statement of “hot topics” like human capital management and other ESG matters that are particularly relevant to them. Human capital is both a board and a management issue, so companies should consider addressing both the board’s (and any committees’) role—in overseeing areas like human capital strategy and corporate culture—and key policies and practices that the company has implemented with respect to topics such as diversity and inclusion, pay equity and employee development. During the last year, there has been an increase in the number of companies that have amended their board committee charters to include explicit language about oversight of human capital and ESG matters.

Companies should also consider highlighting in the proxy statement their other key ESG activities. Summarizing key achievements or targets will assist investors who may not take the time to review more detailed documents such as sustainability or corporate social responsibility reports. Describing company ESG reports available on the company’s website also will alert investors about where they can find more information. For example, in his annual letter to CEOs issued in January 2019,[8] BlackRock CEO Larry Fink asked companies in which BlackRock invests to: (a) publish disclosures in line with industry-specific guidelines issued by the Sustainability Accounting Standards Board by year-end (or disclose a similar set of data in a way that is relevant to their businesses); and (b) disclose climate-related risks in line with the recommendations of the Task Force on Climate-related Financial Disclosures (TCFD), which should include the company’s plan for operating under a scenario where The Paris Agreement’s goal of limiting global warming to less than two degrees is fully realized, as expressed by the TCFD guidelines. Companies that publish these reports or plan to do so in 2020 should consider stating as such in their proxy statements. Such disclosures may also help companies increase their “R-Factor” scores. The “R-” (or “Responsibility”) Factor is State Street’s system for rating companies’ business operations and governance based on financially material and sector-specific ESG issues, using SASB as a foundation. In late January, in a letter to boards of directors,[9] State Street’s CEO announced that in 2020, State Street will begin voting “against” directors at companies in the S&P 500 and certain major non-U.S. indices that are “laggards” based on their “R-Factor” scores and cannot articulate how they plan to improve their scores.

On the governance side, companies may wish to articulate their views on corporate purpose in light of the Business Roundtable’s 2019 Statement on the Purpose of a Corporation[10] and discuss how their practices align with the Investor Stewardship Group’s Corporate Governance Principles for U.S. listed companies.[11] Companies whose major shareholders include State Street should be aware that, as a founding member of the Investor Stewardship Group, State Street “proactively monitor[s]” companies’ adherence to these principles. Consistent with the “comply-or-explain” expectations established by the principles, State Street encourages companies to proactively disclose their level of compliance with the principles. In instances of non-compliance, when companies cannot explain the nuances of their governance structure effectively, either publicly or through engagement, State Street may vote against the independent leader of the company’s board. Finally, companies that have not already started doing so should consider enhancing their disclosures about the board evaluation process, including discussing significant changes that resulted from this process.

At the same time, care needs to be taken when including ESG disclosures in the proxy statement. As with any other public statements, companies should confirm the accuracy of the ESG disclosures they propose to include in the proxy statement. Statements that could be viewed as overstatements or rendered inaccurate by events subsequent to the filing of the proxy statement should be avoided. There should not be significant “divides” between disclosures in the proxy statement and information included in more detailed subject matter publications, such as sustainability or corporate social responsibility reports. Commitments to achieve specific ESG goals or targets, as well as statistics and metrics, should be accompanied by appropriate disclaimer language stating (for example) that goals are aspirational and not guarantees or promises. Companies should think carefully about whether to include hyperlinks to external ESG publications or website content, and it is advisable to include a statement that any hyperlinked material is not incorporated by reference into the proxy statement.

Finally, it may be helpful to include key words to help get credit for ESG practices from various ESG rating services. Governance QualityScore evaluates company practices in four categories: Board Structure, Compensation, Shareholder Rights, and Audit & Risk Oversight. In the compensation category, two new data points now consider the level of disclosure on environmental and social performance measures under short- and long-term incentive plans for executives. On short-term (typically annual) plans, ISS will consider the extent of disclosure of specific performance criteria and disclosed hurdle rates. On long-term plans, ISS will evaluate long-term equity and cash awards granted in the most recent fiscal year based on pre-determined metrics and target goals. For both types of plans, ISS will consider various company-disclosed environmental, social or general sustainability performance measures, such as those focusing on climate change and energy usage or labor conditions in the supply chain.

Don’t Forget to Implement Recent SEC Rule Amendments

In March 2019, the SEC adopted a number of changes to modernize and simplify disclosure requirements.[12] While most of these changes took effect last May, calendar-year filers will be implementing a number of the amendments for the first time in the upcoming Form 10-K and proxy statement. The amendments that impact the Form 10-K are discussed in our recent client alert.[13]

The changes that impact the proxy statement amended Regulation S-K and include:

Revised disclosure heading for late and missed Section 16 reports: The amendments change the disclosure heading required by Item 405(a)(1) of Regulation S-K from “Section 16(a) Beneficial Ownership Reporting Compliance” to “Delinquent Section 16(a) Reports.” Moreover, the amendments add an instruction encouraging companies to exclude this disclosure item altogether if there are no reportable missed or late Section 16 filings.
Simplification of Section 16 due diligence: The amendments also eliminate the requirement for Section 16 persons to furnish reports to companies and clarify that companies may, but are not required to, rely only on Section 16 reports that have been filed on EDGAR (as well as any written representations from the reporting persons) to assess Section 16 delinquencies. This change will likely have limited practical impact, because most companies prepare and file these reports on behalf of Section 16 insiders.
Executive officer information: The amendments revise Item 401 to clarify that any disclosure about executive officers required by Item 401, including information about business experience under Item 401(e), need not be repeated in the proxy statement if it appears in the Form 10-K. This change formalizes an existing staff interpretation on the instruction in Item 401 that permits inclusion of executive officer information in the Form 10-K. Prior to the amendment, the instruction did not explicitly cover all of the information in Item 401.
Update to outdated auditing standard reference: The amendments also update a provision related to the audit committee report by removing a reference to AU section 380, Communication with Audit Committees, which was superseded by Auditing Standard No. 1301. As amended, Item 407(d)(3)(i)(B) of Regulation S-K now includes a general reference to “the applicable requirements of the Public Company Accounting Oversight Board (PCAOB) and the [SEC].” Many companies have already updated their audit committee reports accordingly, but it is advisable to review the audit committee report and confirm the correct auditing standard is referenced or the more general language is used.

Consider Plain English Enhancements

Unlike Forms 10-K, the proxy statement has evolved to become a shareholder engagement tool. As a result, many large cap companies have transformed their proxy statements from legal documents to disclosures that both address various stakeholder concerns and comply with the various SEC and other requirements. All companies would benefit from considering five key questions in order to enhance the readability of their proxy statements:

Short: Does the proxy statement use more words than needed?

Skimmable: Can the reader get the “gist” from a five-minute scan of the proxy statement (g., bold, active headings break up text and “previews” are included for key sections, such as “Why the Board Believes Its Leadership Structure is Appropriate”)?
Logical: Does the flow of information in the proxy statement reflect the disclosure priorities (g., the most important information is placed up front instead of five pages of Q&A)?
Layered: Does the proxy statement include a summary that effectively highlights key themes and significant changes discussed in the proxy statement?
Audience-focused: Is the proxy statement drafted for the various audiences who will review it (g., retail investors, employees, media, SEC, proxy advisors and institutional investors)?

Review These Important Shareholder Proposal Reminders

Companies should keep in mind the following requirements and best practices related to shareholder proposals included in the proxy statement:

Deadline for company response: The statement in response by the company or the board of directors to any shareholder proposal included in the proxy statement must be provided to the proponent(s) at least 30 calendar days before definitive proxy materials are filed with the SEC.[14]
Disclosure about shareholder proponents: The proxy statement must include either the name, address and shareholdings of each proponent or a statement that the company will provide this information promptly upon receiving an oral or written request.[15] Some companies satisfy this requirement using a hybrid approach where there are numerous co-filers: they include in the proxy statement the required information with respect to the lead proponent and then indicate that the name, address and shareholdings of the co-filers will be provided promptly upon request.
Meeting attendance: State law, not the SEC proxy rules, governs the procedures for attending the annual meeting. The proxy statement should describe the extent to which these procedures must be followed by the proponent or its designated representative in order to be admitted to attend the meeting (for example, if pre-registration and identification are required).
Description of shareholder proposals: A shareholder proposal included in the proxy statement must include the title provided by the proponent. However, a company is not required to use the proponent’s title as the title in the company’s table of contents or on the proxy card. Moreover, it is the company’s responsibility to ensure that the description of the shareholder proposal on the proxy card “clearly identif[ies] and describe[s] the specific action on which shareholders will be asked to vote.”[16]

Consider Recent Criticism of Non-GAAP Disclosures in CD&A

Companies should review any non-GAAP disclosures included in the CD&A in light of recent criticisms about the use of these measures and assess whether a reconciliation to GAAP or other additional disclosures are appropriate for certain of these disclosures.

In April 2019, the Council of Institutional Investors (“CII”) submitted a rulemaking petition to the SEC on the use of non-GAAP financial measures in the CD&A. Instruction 5 to Item 402(b) of Regulation S-K provides that when target levels for non-GAAP performance measures are disclosed in the CD&A, companies are not required to comply with the non-GAAP measure disclosure rules in Regulation G and Item 10(e) of Regulation S-K. However, companies still must disclose how the non-GAAP measure is derived from the company’s audited financial statements. In its rulemaking petition, CII asked the SEC to eliminate Instruction 5, thereby making all non-GAAP financial measures in the CD&A subject to Regulation G and Item 10(e), and require that the reconciliation to GAAP be included in the proxy statement or via a hyperlink in the CD&A.

The CII petition echoed other criticism of the increasing use of non-GAAP measures in CD&As, including an April 2019 Wall Street Journal op-ed co-authored by Robert Pozen and SEC Commission Robert J. Jackson, Jr.[17] Despite the CII rulemaking petition and other public criticism, Division of Corporation Finance Director Hinman recently expressed support for the current treatment of non-GAAP measures in the CD&A. At a conference in December 2019, Director Hinman suggested that the current disclosure requirements, coupled with Staff review and comment on the use of non-GAAP measures in the CD&A, are sufficient to address unclear disclosure and “help people understand the comp disclosure better and give more confidence to investors on that process.”[18]

Although it appears unlikely that the SEC will take action in response to the CII petition, in light of ongoing scrutiny and criticism of the use of non-GAAP measures in the CD&A, companies should continue to be mindful of the limitations of Instruction 5 to Item 402(b) and its disclosure requirements. As noted by Director Hinman, where this disclosure is omitted, the Staff will ask companies to comply with Instruction 5 to Item 402(b) and explain how the non-GAAP financial performance measure target level is calculated based on the company’s financial statements. We expect that investors will also be focused on this disclosure. In early January 2020, many large companies received letters from the Say on Pay Working Group, which includes union pension funds and other investors with over $1 trillion in assets under management and advisement. The letter cites the CII petition and urges these companies “to provide clear disclosure in the CD&A of any adjustments to GAAP performance metrics.” In addition, companies including non-GAAP measures outside the CD&A should exercise care in complying with the non-GAAP rules and related interpretations.[19]

Include Additional Disclosures if Holding a Virtual-Only Annual Meeting

Companies that plan to hold a virtual-only annual meeting should confirm that their proxy statement includes appropriate disclosures about the logistics of the meeting and how shareholders can participate.

In recent years, an increasing number of companies have opted to replace their traditional in-person annual shareholder meetings with virtual-only meetings. Although in-person meetings remain the most common approach, according to a recent study by ISS, 7.7% of Russell 3000 companies held a virtual-only annual meeting during the 2019 proxy season—up from 2.4% in the 2015 proxy season. In response to the growing popularity of virtual-only meetings, Glass Lewis implemented a new policy on virtual-only shareholder meetings that took effect in 2019. Under this policy, Glass Lewis analyzes the company’s disclosure of its virtual meeting procedures and may recommend votes “against” the members of the nominating/corporate governance committee if the company does not provide “effective” disclosure assuring that shareholders will have the same opportunities to participate at the virtual meeting as they would at in-person meetings.

In light of increased scrutiny from Glass Lewis and other stakeholders, companies holding virtual-only meetings during the upcoming proxy season should be mindful of providing effective proxy disclosure regarding the virtual meeting. Examples of effective proxy disclosure include:

a brief explanation of why the company chose to hold a virtual-only meeting—this can be of particular utility for companies holding a virtual-only meeting for the first time;
clear instructions on how shareholders can attend the meeting online;
a description of how shareholders can ask questions during the meeting;
the company’s guidelines on how questions and comments will be recognized and disclosed to meeting participants;
procedures for posting questions and answers on the company’s website as soon as practical after the meeting; and
procedures for how the company will deal with any potential technical issues regarding accessing the virtual meeting, including providing technical support.[20]

_____________________

[1] See Frederic W. Cook & Co., Inc., Company Hedging Policies: Observations from New Proxy Disclosures, available at https://www.fwcook.com/content/documents/Publications/
10-10-19_Company_Hedging_Policies_Observations_from_New_Proxy_Disclosures.pdf.

[2] See Final Rule: Disclosure of Hedging by Employees, Officers and Directors, Section III.B.3, available at https://www.sec.gov/rules/final/2018/33-10593.pdf.

[3] See id., Section III.D.3.

[4] Available at https://www.sec.gov/news/public-statement/statement-role-audit-committees-financial-reporting.

[5] Available at https://securitiesregulationmonitor.com/Lists/Posts/Post.aspx?ID=390.

[6] Their respective voting policies are available here (BlackRock), here (Vanguard), and here (State Street).

[7] See ISS, United States: Director Elections & Governance Shareholder Proposals (2019 Proxy Season Review) (Aug. 23, 2019).

[8] Available at https://www.blackrock.com/corporate/investor-relations/larry-fink-ceo-letter?cid=ppc:CEOLetter:PMS:US:NA&gclid=CjwKCAiAsIDxBRAsEiwAV76N869cGeDhtDX
1ekYstnqdXrT0XllD6nasi7-T25o4HIi-NxnuM1h7iRoCZlkQAvD_BwE&gclsrc=aw.ds.

[9] Available at https://www.ssga.com/us/en/institutional/ic/insights/informing-better-decisions-with-esg.

[10] Available at https://opportunity.businessroundtable.org/wp-content/uploads/2019/08/BRT-Statement-on-the-Purpose-of-a-Corporation-with-Signatures.pdf.

[11] Available at https://isgframework.org/corporate-governance-principles/.

[12] Available at https://www.sec.gov/rules/final/2019/33-10618.pdf.

[13] Available at https://www.gibsondunn.com/considerations-for-preparing-your-2019-form-10-k/.

[14] See Exchange Act Rule 14a-8(m)(3).

[15] See Exchange Act Rule 14a-8(l)(1).

[16] See Exchange Act Rule 14a-4(a)(3) and Exchange Act Rule 14a-4(a)(3); C&DI Question 301.01.

[17] See Robert J. Jackson, Jr. and Robert C. Pozen, Executive Pay Needs a Transparent Scorecard, Wall Street Journal (Apr. 10, 2019), available at https://www.wsj.com/articles/executive-pay-needs-a-transparent-scorecard-11554936540.

[18] See Soyoung Ho, SEC Official Signals No Rulemaking on Non-GAAP Measures for CEO Pay, Thomson Reuters (Dec. 18, 2019), available at https://tax.thomsonreuters.com/news/sec-official-signals-no-rulemaking-on-non-gaap-measures-for-ceo-pay/.

[19] In particular, C&DI 108.01 (Non-GAAP Financial Measures) addresses non-GAAP financial information that is not related to target levels and is included in the CD&A or elsewhere in the proxy statement.

[20] For more information, see Broadridge, Principles and Best Practices for Virtual Annual Shareowner Meetings (2018), available at https://www.broadridge.com/_assets/pdf/broadridge-vasm-guide.pdf.

Gibson Dunn’s lawyers are available to assist with any questions you may have regarding these issues. To learn more about these issues, please contact the Gibson Dunn lawyer with whom you usually work in the Securities Regulation and Corporate Governance and Executive Compensation and Employee Benefits practice groups, or any of the following practice leaders and members:

Securities Regulation and Corporate Governance Group:
Elizabeth Ising – Washington, D.C. (+1 202-955-8287, [email protected])
James J. Moloney – Orange County, CA (+ 949-451-4343, [email protected])
Lori Zyskowski – New York (+1 212-351-2309, [email protected])
Brian J. Lane – Washington, D.C. (+1 202-887-3646, [email protected])
Ronald O. Mueller – Washington, D.C. (+1 202-955-8671, [email protected])
Michael A. Titera – Orange County, CA (+1 949-451-4365, [email protected])
Gillian McPhee – Washington, D.C. (+1 202-955-8201, [email protected])
Courtney C. Haseley – Washington, D.C. (+1 202-955-8213, [email protected])
Aaron K. Briggs – San Francisco (+1 415-393-8297, [email protected])
Julia Lapitskaya – New York (+1 212-351-2354, [email protected])

Executive Compensation and Employee Benefits Group:
Stephen W. Fackler – Palo Alto/New York (+1 650-849-5385/+1 212-351-2392, [email protected])
Michael J. Collins – Washington, D.C. (+1 202-887-3551, [email protected])
Sean C. Feller – Los Angeles (+1 310-551-8746, [email protected])
Krista Hanvey – Dallas (+ 214-698-3425, [email protected])

Attorney Advertising: The enclosed materials have been prepared for general informational purposes only and are not intended as legal advice.

1. Mandatory Filings for Critical Technology U.S. Business Transactions

2. Mandatory Filings for Substantial Foreign Government Investments

3. New Exceptions to Mandatory Filings Requirements

4. Declarations for All Transactions—Still No Filing Fees

5. Preliminary List of Excepted Foreign States

6. Clarifying the Excepted Investor Provision

7. Investment Fund Carve-Out

8. New Principal Place of Business Interim Rule

9. Personal Data Collections

10. Real Estate Transactions

Conclusion

Key Definitions Clarified

Notices At Collection Must Be Readily Accessible

“Do Not Sell” Provisions And Optional Opt-Out Button

Responding To Requests To Know And Requests To Delete

Service Providers Are Granted Greater Leeway

Clarifications Regarding Requests To Opt-Out

Recordkeeping Requirements Lessened

Conclusion

I. Filing and Settlement Trends

A. Filing Trends

B. Mix of Cases Filed in 2019

1. Filings by Industry Sector

2. Merger Cases

C. Settlement Trends

II. What to Watch for in the Supreme Court

A. Disgorgement in SEC Enforcement Actions

B. Intersection Between Securities Laws and ERISA

III. Delaware Developments

A. The Delaware Court of Chancery Continues to Scrutinize Relationships Between Directors

B. Court of Chancery Interprets Demand Letter

C. Despite Akorn, an MAE Is Still a Rare Event Requiring a Buyer to Carry a Heavy Burden

IV. Lower Courts Grappling with Implications of Lorenzo

V. Falsity of Opinions – Omnicare Update

VI. Halliburton II Market Efficiency and “Price Impact” Cases

VII. Statute of Limitations for Martin Act Claims Extended to Six Years

I. Statutory Control Under the Bank Holding Company Act

II. New Presumption of Non-Control – 9.99% Rather Than 4.99% Voting Shares

A. Significantly Expanded Rights for Investors at 4.99% Voting Shares or Less

B. Presumptions of Control

C. “Management Agreements” Versus “Rights That Significantly Restrict Discretion”

III. Federal Reserve Math for Calculating Percentage of Voting Shares

IV. Restrictions on Non-Voting Shares

V. Federal Reserve Math for Calculating Total Equity

VI. Special Rule for Investment Funds

VII. Divestiture of Control

Conclusion

APPENDIX 1

TABLE OF CONTENTS

I. GLOBAL POLICY DEVELOPMENTS

A. OECD Recommendations on AIB. Global Partnership on AIC. Global Initiative on Ethics of Autonomous and Intelligent SystemsD. UN Considers Ban on Lethal Autonomous Weapons

II. U.S. NATIONAL POLICY & KEY LEGISLATIVE EFFORTS

A. U.S. National Policy on AI Takes ShapeB. U.S. Imposes Export Controls Related to AI

III. EU POLICY & REGULATORY DEVELOPMENTS

A. EC Focus on Comprehensive AI RegulationB. Ethics Guidelines for Trustworthy AIC. German Data Ethics Commission Report

IV. REGULATION OF AI TECHNOLOGIES AND ALGORITHMS

A. Algorithmic AccountabilityB. Facial Recognition SoftwareC. Deepfake TechnologyD. Autonomous VehiclesE. Data PrivacyF. Intellectual PropertyG. Law EnforcementH. Health CareI. Financial ServicesJ. Labor and Hiring

I. GLOBAL POLICY DEVELOPMENTS

A. OECD Recommendations on AI

B. Global Partnership on AI

C. Global Initiative on Ethics of Autonomous and Intelligent Systems

D. UN Considers Ban on Lethal Autonomous Weapons

II. U.S. NATIONAL POLICY & KEY LEGISLATIVE EFFORTS

A. U.S. National Policy on AI Takes Shape

1. Executive Order “Maintaining American Leadership in Artificial Intelligence”

2. The GrAITR Act (H.R. 2202)

3. Artificial Intelligence Initiative Act (S. 1558)

4. AI in Government Act (H.R. 2575/S. 3502)

5. OMB Guidance for Federal Regulatory Agencies

6. National Security and Military Use

a) The National Artificial Intelligence Research and Development Strategic Plan: 2019 Update

b) NSCAI Report on U.S. National Security

B. U.S. Imposes Export Controls Related to AI

III. EU POLICY & REGULATORY DEVELOPMENTS

A. EC Focus on Comprehensive AI Regulation

B. Ethics Guidelines for Trustworthy AI

C. German Data Ethics Commission Report

IV. REGULATION OF AI TECHNOLOGIES AND ALGORITHMS

A. Algorithmic Accountability

1. H.R. 153

**C. Despite Akorn, an MAE Is Still a Rare Event Requiring a Buyer to Carry a Heavy Burden**

A. OECD Recommendations on AI
B. Global Partnership on AI
C. Global Initiative on Ethics of Autonomous and Intelligent Systems
D. UN Considers Ban on Lethal Autonomous Weapons

A. U.S. National Policy on AI Takes Shape
B. U.S. Imposes Export Controls Related to AI

A. EC Focus on Comprehensive AI Regulation
B. Ethics Guidelines for Trustworthy AI
C. German Data Ethics Commission Report

A. Algorithmic Accountability
B. Facial Recognition Software
C. Deepfake Technology
D. Autonomous Vehicles
E. Data Privacy
F. Intellectual Property
G. Law Enforcement
H. Health Care
I. Financial Services
J. Labor and Hiring