The term “Artificial Intelligence” is not helpful to our public discourse. Artificial Intelligence (AI) is not intelligent. The term encompasses too much, is poorly defined, and therefore can’t be discussed precisely.
But it is important for policymakers to understand what they are encouraging or prohibiting. Passing a law to “restrict artificial intelligence” is a dangerous exercise under current definitions.
Different functions of artificial intelligence create different problems for law and society. Generative AI creates not only new text, code, audio, or video, but problems with deepfakes, plagiarism, and falsehoods presented as convincing facts. AI that predicts whether a prisoner is likely to commit future crimes raises issues of bias, fairness, and transparency. AI operating multi-ton vehicles on the road creates physical risks to human bodies. AI that masters the game of chess may not raise any societal issues at all. So why would politicians and courts treat them the same?
They shouldn’t, but if people don’t understand the distinctions between functional types of artificial intelligence, then they won’t be able to make sensible rules. Our language is holding us back. We need to think differently about AI before determining how to treat it.
There can be useful arguments for defining artificial intelligence by the process used to create it. Large language models or other models built by shoveling tons of data into the maw of a machine-learning algorithm may be the purest form of AI. Politicians don’t understand the distinction between these models and other functioning forms of code, however, and they shouldn’t need to. Politicians don’t care how to build an AI model; they only care what it does to (or for) their constituents.
Some of what we think of as AI is nothing more than complex versions of traditional computational algorithms. Standard big-data mining can seem miraculous, but no machine-learning modules are needed to elicit the desired results. And yet, when regulators discuss strapping restrictive rules onto AI, they would include standard algorithms.
Science fiction writer Ted Chiang has defined artificial intelligence as “a poor choice of words in 1954,” preferring instead to call our current technologies “applied statistics.” He also observed that humanized language for computer activities misleads our thinking about amazing, but deeply limited tools, like effective weather predictors and art generators. There is sorting, excluding, selecting, and predicting in these applied statistical processes, but not context, thinking, or intelligence in the human sense.
Whether our problem is understandable-but-unfortunate humanization of these models, whether it is imprecise thinking about what types of technology constitute AI, or whether it is lumping together of disparate functionalities into a single unmanageable term, we are harming the discourse—and our ability to diagnose and treat dysfunction—by using the term “artificial intelligence” in the present manner.
If we wish to police AI, our society needs to define and discuss AI precisely.
In the explosion of commentary surrounding generative AI, hand-wringing about singularities devolved into an oft-expressed desire to regulate and otherwise “build guardrails” for AI. Society’s protectors, elected and otherwise, believe that we must stop AI before AI stops us, or at least before our use of AI foments foreseeable harm to populations of innocents.
What we casually call AI right now is a set of computerized and database-driven functionalities that should not be considered—and certainly should not be regulated—as a single unit with a single rule. AI consists of too many tools raising too many separate and unrelated societal problems. Instead, if we wish to effectively legislate AI, we should break the definition into functional categories that raise similar issues for the people affected by the technology in that category.
I propose a modest organizational scheme below to assist lawyers, judges, legislators, and regulators to 1) grasp the present state of AI and 2) design rules to regulate the functions of machine learning modules. Some of these lines blur, and certain technical or social problems are shared across classifications, but thinking of current AI solutions in legally significant functional categories will simplify effective rulemaking.
Each of these categories provides a unique set of problems. Legislators and regulators should be thinking of AI in the following functions.
Automating AI: Certain AI models automate processes within business or government without making decisions about the opportunities of specific people. Automating AI may streamline an accounting system, research the case law on burglary, or provide a system of forms for running a company. It can replace human workers in some tasks, and therefore has a social impact.
Generative AI: By predicting the output requested from a series of prompts, certain AI tools build a word-by-word or pixel-by-pixel product that can mimic (or copy) human-looking creations. This can lead to working software code and functioning websites, art in the vernacular of Jan van Eyck, papers discussing the use of symbolism in The Scarlet Letter, or legal arguments in a contract litigation. These products raise intellectual property and plagiarism questions, and they can be trained on improperly obtained material. This technology can produce deep-fakes that are indistinguishable from actionable proof. When Generative AI works poorly, it can generate absolute nonsense presented as true fact.
Physical Action AI: Driverless vehicles operate on the interplay between sensors and predictive algorithms, and so do many industrial and consumer technologies. These are systems that use AI to function in the physical world. This category of AI can include running a single machine, like a taxi, or managing traffic systems for millions of vehicles or Internet of Things devices. Legal concerns not only involve safe product design and manufacture, but tort law, insurance issues, and blame-shifting contracts that affect all activity where the laws of physics and moving bodies apply.
Strategizing AI: We all know about predictive machine-learning programs that mastered human strategy games like chess by playing millions of games. Strategizing AI makes predictions based on running simulations, and humans use it to choose effective strategies related to those simulations. Strategizing AI raises concerns regarding accuracy and reliability of the predictions.
Decisioning AI: This category is not defined by technology but by the technology’s effect on people. Algorithmic tools are used to limit or expand the options available to people. AI ranks resumes for human resource managers, highlights who should be interviewed for jobs, and evaluates the reactions of applicants in those interviews. For years algorithms have made prison parole recommendations, sorted loan applicants, and denied suspicious credit transactions. These decisions are subject to bias of various types, also raising issues of transparency, accuracy, and reliability. The EU and some US states have already regulated this category of AI. The Chinese government has elevated these tools into a societal scoring system that can affect every aspect of a citizen’s life and freedom.
Personal Identification AI: Also defined by its effects on humans, certain AI is being used to pick an individual out of a crowd and name them, to extrapolate whose fingerprint was pulled from a crime scene, or to identify a person carrying a specific phone over specific geography. Most Personal Identification AI uses biometric readings from face, voice, gait, or other traits tied to our bodies, but some is behavioral, including geolocation patterns, handwriting, and typing. This type of AI has been implicated in constitutional search and seizure issues, highlighted for findings that were biased against certain ethnic groups, and questioned for its trustworthiness.
Differentiating AI (Data Analytics): Some algorithms simply decide which items should be included or excluded from a specific group. While this sounds like a simple task, sheer numbers and/or complexity can make the work impossible for humans. Differentiating AI can spot a growing cancer from a shadow on an X-ray much more effectively than teams of trained radiologists. It can predict which cell in a storm may drop a tornado. It helps decide which picture shows a cat and which shows a dog. Both Decisioning AI and Personal Identification AI are legally significant subgroups of Differentiating AI, but I am proposing that this classification will not include tools designed to identify people or make subjective decisions about people, but instead be limited to those that suggest factual groupings that might lead to real-world consequences.
Military AI: An amalgam of each of the other functional types listed here, the military builds its own strategizing models, decisioning models, and physical action AI. These tools raise some of the same issues as civilian versions, but an overlay of special purposes and the law of war create a unique category of considerations for military AI. Like most military tools and strategy, the rules for military AI will appear in international treaties and informal agreements between nation-states. Civilian authorities are unlikely to develop effective limitations for the military’s utilization of AI and algorithmic tools.
AI exists in extensive forms and functionalities, so attempting to regulate the entire set of technologies would be overreaching and likely ineffective. The above categorizations provide a safer place to start if we wish to regulate a vast and shifting technology. By adopting this thinking, AI management becomes less daunting and more effective.
This article is excerpted from To a High Court: Five Bold Law Students Challenge Corporate Greed and Change the Law by Neil Thomas Proto (FriesenPress, 2023). Proto’s first-person account chronicles the efforts of a group of law students—chaired by Proto at the time—to challenge an increase granted by the Interstate Commerce Commission to railroad freight rates nationwide, suing the U.S. and the ICC for violating the National Environmental Policy Act. Their lawsuit culminated in the landmark 1973 decision on Article III standing United States v. Students Challenging Regulatory Agency Procedures (SCRAP).
The Supreme Court, February 28, 1973
The Chief Justice calls the first case: “TheAtchison, Topeka and Santa Fe Railroad Company versus the Wichita Board of Trade, Number 72214.” The Court is being asked to review an order of the Interstate Commerce Commission. The order permits particular railroads to increase their charge to grain shippers by 100 percent to stop the train in transit, inspect the grain, and determine its grade. The agrarians object. They are the railroads’ captives. It was the rough and quixotic abuse of this dependence that spurred the Granger movement and the presidential efforts of William Jennings Bryan. It also moved literary realists like Theodore Dreiser and Frank Norris— abhorred, angry, seeking to further awaken the nation to the railroads’ intimidating grip on the land. Norris said it: “[T]he symbol of a vast power … leaving blood and destruction in its path; the Leviathan, with tentacles of steel clutching into the soil, the soulless force, the iron-hearted power, the monster, the Colossus, the Octopus.” In 1887, Congress created a national institution to regulate the railroads and to protect the public interest: the Interstate Commerce Commission. That attempt had not worked. Montana’s Senator Mike Mansfield sought yearly to legislate the Commission’s abolition. He could not do it.
The Chief Justice thanks the participants in a pro forma but polite way. He settles back into his chair. The attorneys representing the next set of adversaries move into place. George places his hand firmly around my forearm. “We got here,” he whispers evenly. John looks at the two of us and smiles. It is one of his cunning grins. He sits back, prepared to listen. He will not miss a word. Neither will I.
“We will hear argument next in 72535 and 562, United States and ICC against Students,” says Warren Burger, “and Aberdeen and Rockfish Railroad against Students.” The Chief Justice pushes his hair back, seeming to bring recognition to its whiteness. He adjusts his posture—erect, framed, subtly moving his head downward so as to assure himself that the attorneys for each side are prepared to argue their case. He obviously has done this before.
The solicitor general of the United States, Erwin Griswold, has taken his seat at the counsel table directly in front of the Chief Justice and slightly to his left. It is Griswold’s responsibility to represent the United States and the Interstate Commerce Commission. He is the former dean of the Harvard Law School and is experienced in Supreme Court argument. He is wearing his morning coat. Seated next to him is Hugh B. Cox, senior partner at Covington & Burling, Washington’s most prestigious law firm. He is representing the nation’s railroads. Cox also is part of the tradition. He is familiar to this setting and knowledgeable about its culture. He, too, is wearing a morning coat.
Seated at the other end of the table is the attorney for the appellee, Students Challenging Regulatory Agency Procedures (SCRAP). Peter Harwood Meyers is twenty-six years old and a 1971 graduate from the George Washington University School of Law. This is the third oral argument he has ever presented. The first two were in the SCRAP case in the court below.
Meyers is sitting erect, hands folded on the table, nodding slightly when the Chief Justice looks downward. Meyers has trimmed his hair to a fashionable length and is wearing a white shirt. While he has no morning coat, he looks comparatively better without the lavender shirt he wore in the court below. Also at the table, to Peter’s left, is Professor John Banzhaf. “B as in Boy, AN, Z as in Zebra, HAF.” That is the way he spells it for the press, an institution he is acutely sensitive to and depends on for success. He, too, is without a morning coat. Despite the respectful appearance of the two, I know for a fact they have not done this before.
We had decided to file this lawsuit against the Interstate Commerce Commission in May 1972 because the Commission had failed to prepare a detailed statement on the environmental impact of a rate increase granted to all the nation’s railroads on all freight. This obligation, we contended repeatedly to the Commission since fall 1971, was imposed on the Commission by a new law, the National Environmental Policy Act (NEPA). The increased rates are discriminatory. They encourage the extraction of natural resources, including iron ore and timber, and they discourage industrial use of recyclable materials such as scrap iron and steel and textile and paper waste. They also impede the ability of the nation’s cities to move enormous amounts of solid waste. We urged the Commission to consider the environmental consequences nationwide, the same geographic reach of the railroads’ own rate increase.
The Commission and the railroads disagreed. It is their custom to do so. The 1970 report by Ralph Nader’s study group characterized the Commission as a place where “the men in … upper staff … share a protective attitude to ward the transportation industry. They are afraid of change.” This custom of insulation has a legal counterpart. The Supreme Court has regularly deferred to the Commission’s decisions. The Commissioners can do largely what they have always done: resist the public and protect the railroads. They knew that fact going into our lawsuit. So did we.
SCRAP is holding its own. In an opinion written in July 1972, a specially impaneled three-judge district court agreed with SCRAP’s position. “[SCRAP] alleges,” Judge J. Skelly Wright wrote, “that this price increase will discourage the environmentally desirable use of recyclable goods and that … under the terms of NEPA, cannot take effect before a ‘detailed statement on the environmental impact of the proposed action’ is prepared.” The court stopped the rates from going into effect. The railroads and the Commission are uneasy; their relationship has been penetrated, restricted by law, pronounced by a court of law—at the request of five law students. Since the lower court’s decision, the nation’s railroads have been stopped from collecting approximately $400,000 a month for one recyclable material alone: iron and steel scrap. We got their attention. The railroads, along with the United States and the Commission, immediately appealed to the Supreme Court. It is the first time the full Court will interpret the meaning of NEPA. The Chief Justice has already made his views known. He is for the Commission.
There is, however, a threshold legal question. It stems directly from the Constitution. In Article III, the framers created the Supreme Court and limited its jurisdiction to deciding only “cases … and … controversies.” Any party that attempts to invoke the Court’s jurisdiction has to show that it has, in fact, been injured by the government’s action. A party able to make that showing is said to have “standing to sue.” It is not done easily. The Constitution does not tell you what is required. The specific requirements are set by the Supreme Court speaking through its decisions. The requirement has changed over time. Without “standing to sue,” the Court must dismiss the lawsuit.
In the district court, the United States and the Commission challenged SCRAP’s “standing.” In its brief, the government made its case directly: “The plaintiff’s [SCRAP’s] lack of standing to maintain this action was settled by the recent decision of the Supreme Court in Sierra Club v. Morton,” rendered only a few weeks before our lawsuit was filed. According to the government, SCRAP has an even lesser claim: “The complainant in that case, Sierra Club, was a significant and respected organization, with a longstanding concern for the preservation of the environment…. S.C.R.A.P.’s purpose may be commendable, and the zeal of the five law students who comprise it perhaps should be encouraged. However, we respectfully submit that its concern for the public wellbeing does not give S.C.R.A.P. the requisite standing to maintain this action.”
The district court rejected the government’s argument. It concluded that SCRAP had “standing to sue.” But the question is not settled. The United States and the Commission, joined by the nation’s railroads, continue to raise it before the Supreme Court. If the members of SCRAP lack “standing to sue,” then the Commission’s failure to comply with NEPA cannot be reviewed by a court of law.
The solicitor general rises from the counsel table and steps with authoritative comfort to the podium directly in front of the Chief Justice. Solicitor General Griswold’s head moves slightly left to right, seeming to scan the bench as if acknowledging that everyone is present. Hugh Cox is seated to Griswold’s right. Cox is prepared, formidable in skill, and potent in the history and industrial force that buttress his position.
“I am representing the United States and the Interstate Commerce Commission,” the solicitor general states. “Mr. Cox is representing the appellant railroads in No. 72562. We have filed separate briefs, but there is no divergence between our positions.”
Finally, it is beginning. The line is drawn clearly. Now we will see whether it makes a difference that they have been here before and we have not.
The recent US Federal Trade Commission (“FTC”) proposal to ban employers from imposing non-competition clauses (“non-competes”) on their workers has sparked lively debate. Support comes from those who believe the proposed rule would encourage entrepreneurship and innovation and ultimately accelerate economic growth.[1] On the other hand, critics of the proposed rule argue that non-competes are necessary to protect confidential information, such as marketing strategies or pricing plans,[2] although much of the opposition is focused on whether the FTC has the legal authority to pass the rule.
It is difficult to predict whether the proposed rule will actually come to pass, but it received close to twenty-seven thousand public comments. It may be instructive to consider the treatment of non-competes in other common law jurisdictions like the United Kingdom (“UK”) and Singapore, where competing interests between the right of an employer to protect its confidential information, retain employees, and reduce competition are balanced against employees’ rights to market mobility, sometimes in slightly different ways and with different outcomes.
This article will attempt to briefly summarize the employment landscape in each jurisdiction with respect to non-compete agreements; highlight some differences between them; and in turn demonstrate the nuances between the jurisdictions, and their reputations as pro-employer or pro-employee jurisdictions.
I: Overview of New York’s Employment Landscape
In New York, restrictive covenants are “disfavored.”[3] However, they are generally enforceable if narrowly tailored to protect legitimate business interests such as trade secrets, confidential information, or customer goodwill and if they do not unreasonably restrict an employee’s right to earn a living. The duration and geographical scope of the non-compete must also be “reasonably limited.”
In the seminal case of BDO Seidman v. Hirshber, [4]the New York Court of Appeals held that a non-compete agreement that prevented an accountant from working for any client of his former employer for eighteen months, even if he had never worked on that client’s account, was overly broad and unenforceable. The court noted that such a restriction would effectively prevent the employee from working in his chosen profession, and was unnecessary to protect the employer’s legitimate interests.
The FTC proposal seeks to place a broad ban on non-competes, claiming that non-competes affect more than thirty million people in the private sector and if banned, would increase American workers’ earnings by $250 to $296 billion.
It is submitted that the new FTC rule would arguably produce limited practical changes in litigation outcomes in New York, whereas the potential practical impact of the rule in other arguably more pro-employer states like Maryland could be more significant.[5]
II: Overview of the UK’s Employment Landscape
Across the Atlantic, the enforceability of non-compete agreements is subject to common law principles and statutory regulations. Generally, non-compete clauses are enforceable only to the extent that they are reasonable and necessary to protect the employer’s legitimate business interests. The case of Tillman v Egon Zehnder Ltd[6] clarified the law on restrictive covenants and represents a good illustration of how the UK courts balance competing interests in this area. There, the Supreme Court held that the non-compete clause was unenforceable as it was too wide and prevented Tillman from holding even a minor shareholding in a competing business.
In Marathon Asset Management LLP v Seddon,[7]Seddon was sued for breaching his six-month non-compete clause. The Court of Appeal held that the non-compete clause was too wide and therefore unenforceable, as it prevented Seddon from working for any competing company, regardless of whether it was a direct competitor.
These cases demonstrate the UK’s increasingly strict approach to non-competes, where reasonableness and necessity are scrutinized and overly broad or restrictive clauses can be deemed unenforceable by the courts.
III: Overview of Singapore’s Employment Landscape
In Singapore, non-compete agreements are enforceable to the extent that they protect an employer’s legitimate business interests and are reasonable in scope, duration, and geographical coverage. Singapore is known as an employer-friendly jurisdiction, but its courts do strike down clauses that they deem overly restrictive or unreasonable. Indeed, in some respects, Singapore may be said to police the reasonableness of certain aspects of non-competes, for the benefit of employees, more strictly than New York.
For example, New York courts generally regard the use of non-competes to protect confidential information as legitimate. However, in Singapore, if the protection of confidential information is already addressed by another clause in the contract (e.g., a confidentiality clause), the Singapore courts have held that the employer must demonstrate that the non-compete clause covers a legitimate proprietary interest over and above the protection of confidential information or trade secrets, in order not to be regarded as an unenforceable restraint of trade.[8]
IV. Same but different
While the legal tests in each jurisdiction are very similar, their application to broadly similar facts can sometimes result in divergent outcomes, reflecting the public policy considerations of the respective jurisdictions. In the UK, courts have taken a relatively strict approach to the enforceability of non-compete clauses due to the country’s public policy of promoting competition and preventing the restriction of trade. As such, non-compete clauses must be narrowly tailored to protect the legitimate interests of the employer without unduly restricting the employee’s ability to work.
In Singapore, non-compete agreements are generally viewed as a legitimate means for businesses to protect their proprietary interests, within certain boundaries related to the duration, scope, and nature of the restrictions.
Given the substantial and often negative influence of large corporations on “labor market fluidity” in the US, with non-competes common even for fast-food workers, clerks, and low-level hospital employees, the FTC’s proposed rule should be welcomed by workers. Whether or not the rule is implemented, employers in the US would be well advised to start exploring alternative contractual, technological, and practical mechanisms to protect their confidential information, trade secrets, and goodwill.
This article originally appeared in International Law News, the quarterly magazine of the ABA International Law Section, in the Spring 2023 issue (Volume 50, Issue 3). Join the International Law Section to read the full issue and access other resources regarding international law.
Title 3 of Maryland’s Code Annotated for Labor and Employment, Section 3-716 only prohibits employers from requiring non-competes when the employee is earning equal to or less than $15 per hour. ↑
Today, some legal experts argue it’s time for Big Law to get more personal. Rather than prioritizing long work weeks and high salaries, firms need to focus on efficiently and carefully meeting their clients’ needs.
That’s where boutique law firms come in. These firms, typically consisting of fewer than thirty experienced lawyers, offer tailored services and have shorter, more intentional client lists.
In many ways, boutique law firms are disrupting the legal industry and paving the way for future innovations. Let’s dive deeper into their benefits and impact on the industry.
(1) Personalization
Boutique firms have a narrower focus and take on a smaller number of cases. This can allow them to deeply engage with each client and provide more custom attention to their needs.
According to research from Law Firm Marketing Club, 84% of clients expect same-day responses to queries. Another 65% expect to be able to speak to an attorney themselves, and 60% expect to have online chat options with their firm.
The bottom line: regardless of the legal subject matter, clients want to have a personal relationship with a responsive attorney—and boutique law firms have the opportunity to give that to them in a different way.
Bigger firms tend to have bigger deals, and with those come greater responsibility and more administrative staff. This can be a con in some situations, as mid-level or junior associates may take on more of the workload, resulting in less face-to-face time between clients and partners.
As with any small business, boutique firms have the opportunity to communicate more directly with clients. They are not bogged down by administrative demands of larger firms, which can result in quicker response times and more personal conversations.
(2) Expertise and Agility
Boutique law firms offer more dynamic representation for clients by leveraging knowledge and expertise in a particular area of law. When clients hire a smaller firm, they are usually opting for expertise over strength and size—and that can pay off.
Many attorneys choose to work at a boutique firm rather than a large firm because they’re passionate about a particular practice. A close-knit specialty environment allows lawyers to zero in on what they’re best at (and most interested in), resulting in more targeted services.
Boutique firms also tend to be agile and adaptable when responding to both client requests and industry regulations. Their small teams, coupled with their high degree of specialty and personal client attention, allow them to adjust to relevant legal changes and trends quickly.
(3) Operations
Finally, boutique law firms may offer effective, direct representation through streamlined operations that leave more room for client services.
Considering that most small firms boast a flatter organizational structure, there are fewer layers of management, reducing the risk of bureaucracy. At the same time, they benefit from having enough team members to get tasks done with urgency and precision.
This culture doesn’t just benefit the client – it can also have a significant impact on the attorneys’ work-life balance. These attorneys will likely get more hands-on experience and client-facing time, but they also tend to have more manageable caseloads and fewer 10+ hour work days.
Conclusion
As the world has become increasingly focused on personalization and speed, it’s time for the legal field to catch up. Boutique firms can help push the field in that direction.
Deciding between a boutique firm and a larger firm depends heavily on your preferences and case-specific needs. The bottom line is that you have options, and evaluating them is the first step in selecting the right partner.
The Ukraine conflict is the most recent event in a long history of strife between Russia and Ukraine that predates the current invasion by 250 years.[2] In relevant recent events, the Maidan Revolution in November 2013 was marked by mass protest against the Ukrainian government’s decision to cease negotiations with the European Union (E.U.) to enter into an association agreement, which was blamed on Russian influence.[3] After 77 protestors were killed in Kyiv, then-President Viktor Yanukovych was exiled to Russia while the Ukrainian opposition party took control of the government.[4] In March 2014, Russia annexed the Crimea, and in May 2014 pro-Western politician Petro Poroshenko won the Ukrainian presidential election.[5] In April 2019, Volodymyr Zelensky became president, unseating incumbent Poroshenko.[6] On February 21, 2022, Russia recognized two breakaway Ukrainian territories, Donetsk and Luhansk, which Russia had been supporting as independent states, and sent significantly more military support into the territories.[7] On February 24, 2022, Russia invaded Ukraine.[8]
The United States has had economic sanctions in place against Russia since 2014, when it invaded Crimea. However, the reaction to Russia’s wholesale military incursion in 2022 has taken the extent and degree of sanctions to a far higher level, creating the need for U.S. companies to establish much stronger diligence programs and in some cases deal with complex compliance issues.[9] The adoption of varying sanctions by numerous other countries and Russia’s recent moves to impose countersanctions and take steps against companies complying with international sanctions have made operating in the global marketplace a complicated and high-risk proposition.[10] Russian countersanctions have focused on indirect expropriation as leverage to punish or control foreign investment in its borders, using the countersanctions framework to extract assets and value from foreign investors. The current countersanctions regime is being implemented in a manner reminiscent of the Russian government’s long-standing use of bankruptcy and taxation law to punish corporations that fell out of favor or ran into conflict with the government.
Russian Bankruptcy Laws
Russian insolvency matters are governed by Federal Law No. 127-FZ On Insolvency (Bankruptcy) (the “Russian Bankruptcy Law”), which has been amended repeatedly since its introduction on October 26, 2002.[11] In late December 2008, further reforms were enacted, including provisions on bankruptcy administrators, payment priorities, secured creditors’ rights, foreclosure on security during bankruptcies, and settlement of taxes.[12] The impact of the Russian Bankruptcy Law and its subsequent amendments has evolved over time; after a few years, it created concern among businesses operating in Russia that the laws would be politically misused because politicians had too much of a role in the process.[13] The implementation and execution of the Russian Bankruptcy Law by the Russian government has drawn criticism; such criticism and concern has become acute and renewed in light of the Ukraine conflict.
Perhaps the most significant development in Russian insolvency law prior to the enactment in 2002 of the Russian Bankruptcy Law occurred with the 1998 legislation, which replaced the 1992 bankruptcy law. The 1992 law was itself a new legal regime instituted in the wake of the fall of the former Soviet Union.[14]
The 1998 amendments were a keen source of interest for both the media and legal scholars because they were heralded as a fundamental reform of bankruptcy laws that had allowed debtors, prior to the amendments, to run up insurmountable debt, which left very little to distribute to creditors or important constituencies like the Russian government and workers.[15] The prior laws were criticized as being too debtor friendly, allowing weaker companies to continue to finance their struggling operations by not paying creditors.[16] Many Russian economic and insolvency experts argued that the bankruptcy laws facilitated heavy payment defaults that Russian debtors were being allowed to accumulate before they were declared bankrupt.[17] The 1998 amendments’ hallmark, and the topic of much of the discussion, was one of hope for a new economic pragmatism towards bankruptcy law that would help modernize the Russian economy.[18] At the time, a payment default rate in Russia existed that threatened the viability of the economy, and those in support of the reforms felt that the bankruptcy laws were too lenient on defaulting debtors.[19]
In April 2009, further amendments were made that included provisions on vicarious, third-party liability in a bankruptcy; lower voluntary bankruptcy filing thresholds; rules regarding conflicts by business entities; and provisions regarding bankruptcy litigation tools such as avoidance actions used to bring assets back into a bankruptcy estate (preference actions or fraudulent conveyances, for example).[20] On July 30, 2017, significant changes were made to the Russian Bankruptcy Law regarding vicarious liability of controlling persons.[21] These and other successive amendments to the Russian Bankruptcy Law were designed to perfect practical application of its provisions and to prevent perceived abuse by shareholders.[22] Because of the prevalence of personal guarantees in Russia, it is also important to note that in 2015, the Russian insolvency laws introduced a personal bankruptcy act.[23]
Russian “Rule of Law”
Though structural improvements have been progressively better through each amendment of the Russian Bankruptcy Law, there are strong opinions over whether or not the laws matter if such laws are not “honestly enforced.”[24]
Legal scholars, journalists, and courts outside of Russia have been consistent in expressing doubts over the legitimacy of the actions of Russian bankruptcy trustees, also known as bankruptcy administrators. For example, the 1998 bankruptcy law amendments were supposed to create a way for creditors to enforce claims, but they have been criticized as instead being used as a new mechanism for insiders to siphon off funds from minority shareholders and creditors; there have been allegations that the fiscal malfeasance has been facilitated by bribes to judges and bankruptcy trustees.[25]
President Boris Yeltsin’s announcement in late 1999 that he would step down as president at the end of 1999 and designate Vladimir Putin as his successor was a surprise to many.[26] President Yeltsin’s regime was tied closely to the rise of oligarchs who had profited from the large-scale privatization of Soviet Union’s economy.[27] Some hoped that this surprise successor would bring a new direction for economic policy focused on further modernization.[28]
The Demise of Yukos
The fate of the Yukos corporation is cited as the definitive indication of how the Putin regime would handle the economy and was lamented as a failure to address the cronyism of the Yeltsin regime.[29] The chairman/CEO of Yukos at the time, Mikhail Khodorkovsky, was a Russian oligarch who attempted to enter into a merger with a Western oil company. The prospective merger was favorably received by the markets with a $43 billion valuation, though the good news came shortly before Khodorkovsky was arrested by the Russian authorities.[30]
The Russian government did not formally privatize Yukos; instead it used the tax laws to orchestrate a takeover that not only jailed Russian nationals, but also rendered foreign and domestic investment worthless through enforcement of the Russian tax code. In April 2001, the Russian government imposed a total assessment of $3.4 billion in taxes owed and sought court enforcement against Yukos, which resulted in an order freezing company assets and forbidding the company from alienating or encumbering its property.[31] The government then engaged in a series of actions that crippled both the company’s finances and its ability to pay any outstanding taxes.[32] A judge who ruled against the government and tried to lift the freeze order was removed from the case and then fired, while a judge who ruled for the government was promoted.[33] The courts also allowed procedural deadlines to be instituted that did not allow for time or access to review documents or evidence.[34] Yukos’s legal department, in turmoil due to the arrests of Yukos personnel, was given exceptionally short deadlines to respond to the government’s case and no effective opportunity to review the government’s evidence.[35] The government also argued that due to the asset freeze, Yukos could not liquidate any assets to pay the fine.[36]
It soon became evident that the true objective of the government was to render the company insolvent and begin taking its assets; this was not done directly, but instead by proxies. In July 2004 the government announced the sale of a Yukos affiliate, YNG, which owned the majority of the Yukos production operations, by selling off the YNG stock owned by Yukos.[37] The government increased the tax liabilities up to $24 billion, and then the sale only yielded $9.35 billion, despite YNG’s stock being valued at between $15 and $20 billion. A state-owned oil company, Rosneft Oil Company, purchased YNG.[38] Yukos filed for bankruptcy protection in 2006; the courts, Rosneft, and the tax authority rejected its restructuring plan, and it liquidated in 2007, with Rosneft taking most of the assets.[39] Promnefstroy, a former Rosneft subsidiary, was given the rights to Yukos’s Dutch subsidiary Yukos Finance B.V.[40]
In March 2022, Russia’s ruling party, United Russia, announced a draft law that provides for the involuntary bankruptcy sale of assets left behind by departing companies from “unfriendly countries.”[41] The proposed law would have placed entities that were 25% or more owned by foreigners from “unfriendly states” into bankruptcy unless they divested their holdings to a Russian entity.[42] While the Russian government placed this draft legislation and similar, more direct, measures aside at that time, a new “Yukos” strategy has emerged out of the battling sanctions and countersanctions that have been implemented because of the Ukraine conflict, and recent countersanctions have involved seizures of foreign companies, which have been justified by the Kremlin as retaliation for seizure of Russian funds.
Russian Countersanctions, Indirect Expropriation
The Russian government has imposed significant “countersanctions” against “unfriendly countries” and controversial third-party sanctions against companies complying with the Western sanctions regimes against Russia and Belarus. This product of the Ukraine conflict was an anticipated, but no less disruptive, byproduct of the U.S., U.K., and E.U. sanctions regimes. The Russian countersanctions have complicated and stymied global efforts by corporations to comply with all applicable laws in the markets in which they operate and has exacerbated the exodus of companies from Russia. The hallmark of these government efforts by the Kremlin and the Russian legislature is a revitalization of the tactic of expropriation—especially indirect expropriation—which harkens back to familiar tactics utilized by the Soviet Union, but adapted to modern global economic and political norms.
Expropriation
Expropriation, the seizure of property of another country’s nationals by a sovereign, is a recognized right of sovereign states (“States”) under certain conditions. For expropriation to be lawful (i.e., where the State does not incur international liability), the taking must be for a State obligation, cannot discriminate against foreigners, must respect due process, and must entail prompt and adequate compensation.[43] If expropriation is legal, compensation may limited “to the value of the company at the time of dispossession, plus interest to the date of payment.”[44] Unlawful expropriation can allow for further damages, including lost profits.[45]
Direct expropriation is a legal transfer of title or the physical seizure of property that benefits the State—or a State-selected third party—accomplished through formal law, decree, or physical act.[46] Large-scale nationalizations are direct expropriation and are rare.[47] Indirect expropriation is the complete or partial deprivation of a property right without a formal transfer of title or seizure.[48] There are many terms for the variants of indirect expropriation, including de facto, creeping, constructive, disguised, consequential, regulatory, or virtual expropriation.[49] Creeping expropriation is worth further examination. Creeping expropriation “encapsulates the situation whereby a series of acts attributable to the State over a period of time culminate in the expropriatory taking of such property.”[50] The State can utilize regulatory, legislative, and judicial processes to interfere with property rights over time to dilute foreign nationals’ rights without transferring title or taking control of the asset.[51]
Modern expropriations typically result from legislative, executive, and administrative acts, which include new legislation; resolutions; decrees; and revocation, cancellation, or denial of government concessions, permits, licenses, or authorizations that are necessary for the operation of a business.[52] Judicial intervention is more rare.[53] Investors have challenged confiscatory tax policy; the prohibition of distribution of dividends; labor regulations or other interference in staffing and operations; judicial decisions; financial regulations; and licensing regimes as expropriation.[54]
Russian Response to Ukraine Conflict
The Russian commission on legislative activities has drafted proposed legislation nationalizing property of foreign organizations leaving the Russian market. It was first announced that the issue would be reviewed by a government commission in March 2022, but the law has not yet been passed.[55] The proposed law on external administration for the management of a company would apply to companies with (1) over 25% shareholding (directly or indirectly) “connected” to “unfriendly” States (including place of registration and place of primary economic activity); and (2) a book value of over one billion rubles (around USD 12 million as of early April 2022) and/or over 100 employees. If passed, this law could have retroactive effect from February 24, 2022.[56]
The proposed legislation would be direct expropriation that would arguably require compensation under international law. Since the Yukos takeover was so successful and was accomplished without payment from the Russian treasury, legislation has typically been geared towards indirect expropriation, which is harder to prove than a direct expropriation and easier to defend against in later lawsuits over compensation.
In May 2021, Russian Prime Minister Mikhail Mishustin signed a decree accompanied by a list of “unfriendly states” that “have carried out unfriendly actions” against Russia, Russian nationals, or Russian entities, primarily by restricting diplomatic relations with Russia.[57] The original designation only included the U.S. and the Czech Republic and was made in response to separate denunciations by the U.S. and the Czech Republic for interference by Russia within the borders of each respective country.[58] The list has been expanded multiple times, and its original purpose has been expanded by the Russian government as a response to sanctions imposed against Russia for its invasion of Ukraine.[59]
On March 5, 2022, the Russian government approved the most recent list of “unfriendly countries,” which included Albania, Andorra, Australia, Canada, members of the European Union, Iceland, Japan, Liechtenstein, Micronesia, Monaco, Montenegro, New Zealand, North Macedonia, Norway, Singapore, San Marino, South Korea, Switzerland, Taiwan (the Republic of China), Ukraine, the United Kingdom (including Jersey, Anguilla, British Virgin Islands, and Gibraltar), and the United States.[60]
Russia also instituted countersanctions by presidential decree on March 3, 2022, including a ban on trade, financial transactions, and the honoring of any contract with any party that is a foreign national of an “unfriendly country,” which is the legal designation for any country that has joined in the sanctions against Russia.[61] The Russian government also passed a resolution that all transactions and operations between Russian companies and nationals from “unfriendly countries” must be approved by the Government Commission on Monitoring Foreign Investment.[62]
Currency Restrictions
The Russian government has also imposed currency restrictions on Russian banks so that their clients cannot withdraw more than the equivalent of $10,000 in U.S. dollars.[63] Otherwise, currency can only be withdrawn in rubles at a rate set by the central bank.[64] Any money deposited since March 9, 2022 cannot be withdrawn. The Russian central bank originally stated these restrictions would remain in place until at least March 9, 2023.[65] The restrictions were recently extended to September 30, 2023.[66]
Patent Licensing
Foreign nationals from “unfriendly countries” are subject to a compulsory license of their patents for a 0% royalty.[67] Owners of Russian patents will receive no compensation for infringement in Russia for as long as the countries remain on the “unfriendly country” list, which is populated with countries that have imposed sanctions on Russia and Russian nationals for Russia’s invasion of Ukraine, as described above.[68] The same type of legislation was extended to trademark and copyright infringement.[69] Hasbro Inc. subsidiary Entertainment One UK, which owns the animated children’s television character “Peppa Pig,” brought a trademark and copyright infringement claim against a Russian citizen in the Russian Arbitration Court for the Kirov Region in 2021.[70] On March 3, 2022 (after sanctions were imposed), the court overturned an award it had given to Entertainment One UK and held that the Russian citizen could continue to use the trademarks without payment or permission.[71] The court ruled that a plaintiff domiciled in an “unfriendly country” cannot claim IP infringement against Russian defendants.[72]
Energy Sector
The most serious restrictions have been in the energy sector, where the government measures have been directly tied to the status of the foreign national’s country on the “unfriendly country” list. For example, there was a ban in 2022 on foreign investment in the energy sector. The law banned foreign investors from unfriendly territories and states from restructuring or selling their interests in Russian strategic enterprises, banks, energy companies, and mining companies until December 31, 2022.[73] In practice this means the government can control the terms under which an investor can restructure/sell its assets and effectively can appropriate the assets indirectly. This law has been applied several times, where foreign businesses were essentially forced to sell or lose their businesses in Russia. The government is able to set the rates centrally under this decree.[74]
In March 2023, Russian countersanctions were expanded.[75] Businesses from “unfriendly countries” are now required to make a voluntary contribution to the Russian state budget, and failure to pay will subject the business to a government-appointed receiver.[76] Shortly after, in April 2023, President Putin authorized expropriation of foreign-owned assets in response to the seizure of Russian assets.[77] The first victims of these policies were, not surprisingly, in the very lucrative energy sector.
In late April, the Kremlin seized the shares of Finland’s Fortum (FORTUM.HE) and Germany’s Uniper (UN01.DE), which operate energy plants in Russia, and placed the shares in the custody of temporary control of Rosimushchestvo, the federal government property agency.[78] Rosneft seized operations and remains in control of the facilities.[79] The Russian government justified the seizures as direct retaliation for seizures of its assets by unfriendly countries.[80]
Consequences for “Unfriendly Countries”
On October 7, 2022, the Russian government moved forward through presidential decree to disenfranchise the owners of the Sakhalin-1 energy project, including the foreign nationals of several countries that the Russian government has designated as “unfriendly countries.” ExxonMobil, a significant owner, operator, and partner in the project, had had difficulty with the Russian authorities since late summer 2022 and initiated a prerequisite to arbitration by sending a “note of difference” to the Russian government after President Putin issued a decree in August 2022 that blocked a sale of its 30% stake in the Sakhalin-1 project.[81] The sale would have effectuated the exit ExxonMobil announced in February 2022 after the invasion of the Ukraine.[82] While not naming a buyer in its SEC quarterly filing, it was reported that Exxon would be selling its stake to existing partner ONGC Videsh of India, which wanted to increase its existing 20% stake.[83] Significantly, India has not joined in Russian sanctions and increased its purchases of Russian oil by 33 times since the invasion of the Ukraine.[84] In April 2022, ExxonMobil disclosed a $3.4 billion write-down on the Russia exit and signaled a third-quarter $600 million impairment charge for unidentified assets.[85] ExxonMobil had valued its Russia holdings at more than $4 billion.
ExxonMobil has denounced the decrees as an expropriation, stating: “With two decrees, the Russian government has unilaterally terminated our interests in Sakhalin-1, and the project has been transferred to a Russian operator.”[86]
Through the October decree, President Putin seized ExxonMobil shares in the oil production joint venture and transferred them to a government-controlled company it established, managed by Rosneft subsidiary Sakhalinmorneftegaz-shelf. Foreign partners will have one month after the new company is created to ask the Russian government for shares in the new entity pursuant to the decree. In addition to ExxonMobil and ONGC Videsh, a foreign-owned entity affiliated with Japan’s SODECO also holds a stake in the project.[87] The decree gives the Russian government authority to decide whether these foreign owners can retain stakes in the project.[88]
President Putin used a similar strategy in a July 2022 decree to seize full control of Sakhalin-2, another gas and oil project in the Russian Far East, with Shell and Japanese companies Mitsui & Co and Mitsubishi as partners.[89] On July 1, 2022, it was reported that Shell had lost its 27.5% stake in the Sakhalin-2 project after the Russian government transferred the project to a new holding company.[90] Japan’s Mitsui & Co and Mitsubishi agreed to take shares in the new holding company, retaining their minority stakes in Sakhalin-2.[91]
Remedies
Multinational corporations have largely been driven out of Russia and suffered a similar fate as Russian citizens who were targeted by the Russian government in the recent past. The Russian government has used regulatory and tax regimes to soften companies, even jail the management, and then seize the assets through the bankruptcy system. The Ukraine conflict has created a sanctions war between the West and Russia where both sanctions regimes have placed corporations in the precarious position of having to navigate between them and risk penalties from lack of compliance within the borders of each regime by compliance in the other.
Another unwelcome product of the conflict and related sanctions regimes is the adaptation of these same takeover techniques to accelerate the exit of foreign companies from Russia and disguise expropriation by using regulatory and legal justifications to seize foreign company assets within Russia. So what recourse is there for foreign companies who have lost millions in their investment and assets in Russia during the Ukraine conflict?
Lessons from the Yukos Bankruptcy
The litigation that followed the dissolution of Yukos provides a useful history regarding the legal issues, opportunities for recovery, and potential liability for those seeking to exercise their creditor’s rights against the Russian government. The Russian government did not stop with the dissolution of Yukos; it pursued employees and management of Yukos who left for fear of criminal prosecution. Extradition requests and subpoenas for records of third-party corporations were denied because the applicable domestic courts outside of Russia had serious reservations about the Russian judicial process, including the criminal and tax proceedings in Russia.[92] The Russian bankruptcy administrator sold Yukos Finance B.V. in the Netherlands and Yukos CIS in Armenia to Rosneft and Promneftstroy, a former Rosneft subsidiary.[93] In assigning the Dutch assets to a locally controlled Dutch trust, the Dutch court held that it was primarily motivated to put the assets in trust by the lack of integrity in the Russian proceedings, which failed almost every possible standard for a proper adjudication of the issues.[94]
U.S. Courts
Bankruptcy courts in the United States are likely to have limited success in untangling insolvency matters in Russia, especially those that are a result of countersanctions. Chapter 15 was added to the U.S. Bankruptcy Code to facilitate cross-border cooperation between U.S. and foreign courts where assets in multijurisdictional disputes are a common challenge.[95] Typical of many treaties, there is a public policy exception in chapter 15, which allows a U.S. bankruptcy court to refuse to cooperate in a cross-border dispute, if cooperation is “manifestly contrary to the public policy of the United States.”[96] The entirety of the current U.S. sanctions regime is based upon the isolation of Russia from the global financial system, and no U.S. bankruptcy court would likely facilitate any payment that would proceed through Russian financial institutions. Section 1503 also prohibits any action that conflicts with a treaty or “other agreement” with other countries.[97] The U.S. currently is acting in cooperation with its allies and partners to impose both export controls and a wide range of sanctions.[98]
Prior to the 2022 invasion of Ukraine, and outside of the current U.S. and E.U. sanctions, a U.S. bankruptcy court granted recognition of a Russian insolvency proceeding as a foreign main proceeding in the In re Vneshprombank case.[99] The U.S. Bankruptcy Court of the Southern District of New York allowed a Russian trustee to take discovery in order to trace and recover a failed Russian bank’s allegedly stolen funds, which were taken by two New York LLCs and deposited within the U.S.[100] Two prior Chapter 15 cases were less successful: one where the Russian foreign representative dropped the matter,[101] and another where provisional relief was granted but the case was dismissed before recognition.[102]
In 2004, embattled Yukos tried to file a U.S. bankruptcy to stop Russian tax authorities from auctioning off its Siberian oil-pumping subsidiary Yuganskneftegaz to collect on the $27.5 billion back tax bill that had been assessed by the Russian government.[103] The U.S. Bankruptcy Court for the Southern District of Texas initially granted a temporary restraining order and preliminary injunction to stop the sale, holding that the evidence presented “support[ed] a finding that it is substantially likely that the assessments and manner of enforcement regarding Plaintiff’s taxes were not conducted in accordance with Russian law.”[104]
The court later granted a motion to dismiss the case brought by Deutsche Bank AG, rejecting arguments made regarding lack of jurisdiction, forum non conveniens, comity, and the act of state doctrine.[105] The court instead dismissed the case based upon a review of 11 U.S.C § 1112(b), holding that a variety of factors warranted dismissal. It held that it was impractical to expect that the Russian government would cooperate with the case (and the court deemed such cooperation essential to the bankruptcy), that funds that were transferred to the U.S. were done so immediately before filing and to create jurisdiction in the U.S., and there were multiple other forums in which Yukos had filed seeking relief which the court felt were as capable as, if not more capable than, the U.S. Bankruptcy Court in determining matters under foreign law.[106]
Multilateral and Bilateral Forums
The U.S. suspended bilateral engagement with the Russian government on most economic issues in response to Russia’s ongoing violations of Ukraine’s sovereignty and territorial integrity.[107] Russia is a party to 60 bilateral investment treaties,[108] including many with countries that are now listed as “unfriendly” under Russian sanctions law.[109] There have been several arbitral awards against Russia related to the annexation of Crimea in 2014 that have had to be enforced against Russian assets outside of Russia.[110] The various measures taken by the Russian government against “unfriendly countries” are in violation of these bilateral investment treaties (and many multilateral treaties).[111] Investment treaties incorporate the basic principles of legal expropriation, which require the taking to have a public purpose, be nondiscriminatory towards foreign nationals, and be subject to due process. Investment treaties also regularly prohibit restrictions on transfers of funds.[112]
Under the Convention on the Recognition and Enforcement of Foreign Arbitral Awards (“New York Convention”), an award against Russia, which is a party to the treaty, can be enforced in the 169 jurisdictions that are also parties to the treaty.[113] There are over 300 billion USD of Russian assets from Russia’s central bank that have been frozen by the U.S. and its allies,[114] though it will take further authority from Congress to attach the assets that have been seized or frozen.[115]
Russia is also a party to the World Trade Organization (WTO), and its measures against the intellectual property of nationals of “unfriendly countries” also could violate the extensive multilateral protections for intellectual property under that umbrella. However, Article XXI of the General Agreement on Tariffs and Trade (GATT) is a “national security exception,” which allows WTO members to breach their WTO obligations for purposes of national security.[116] Article XXI of the GATT was invoked by Russia in a WTO panel dispute between itself and Ukraine over trade restrictive measures taken by Russia, and the Article XXI claim was upheld by a WTO panel on April 5, 2019.[117]
Domestic courts have been largely unsuccessful in suits against the Russian government for lack of jurisdiction.[118] International tribunals also have limited jurisdictional reach. Although Russia was not a party to or had not ratified certain signed treaties, claims have been brought under European Convention of Human Rights in the Strasbourg, though that court has limited remedies available to it.[119]
Between the bilateral tribunals and the Strasbourg Court, litigants against Russia in the Yukos matter fared better under arbitrators.[120] The Russian Chamber of Commerce’s International Commercial Court awarded $425 million to the Dutch trust pursuant to an arbitration provision in the loan documents based on a loan default to which Roseneft became a successor when it took over Yukos assets.[121]
A Spanish arbitrator, pursuant to a bilateral investment treaty, found that unlike the Strasbourg Court, it did not have to analyze whether or not Russia violated its own laws or international norms; instead it had to determine whether nor not Russia engaged in expropriation and acted inconsistently within the expected and normal exercise of regulatory powers to enforce a tax regime.[122] The arbitrator was particularly critical of both the Russian government and the Strasbourg decision under an expropriation analysis.[123] The panel found that Russia had engaged in expropriation, which required compensation.[124]
Conclusion
The accumulating volume and complexity of U.S. sanctions create serious risks for U.S. businesses with respect to both their own vulnerability to sanctions violations and concerns over retaliation from governments involved in the conflict. Proper redress at a later stage for losses incurred due to direct or indirect expropriation will depend on preservation of evidence and strategic planning in anticipation of which forums will be used to recover value from the consequences of the Ukraine conflict.
The pattern that has emerged from Russian countersanctions of using indirect expropriation though taxation, bankruptcy, intellectual property, currency restrictions, and investment controls should be carefully monitored and examined in formulating a strategic approach to the pursuit of recoveries. Equally important is whether assets seized as part of the Western sanctions regime will actually be available to satisfy the losses of these companies once they have found an appropriate forum willing to take jurisdiction over a commercial dispute stemming from a sanctions-induced loss.
The Russian approach to the Western sanctions regime is a direct extension of its recent history of using taxation and regulatory powers to attack real and perceived enemies within its borders, weakening them and then using the bankruptcy laws to take their assets. The countersanction regime justifies the adaptation of this use of government power against foreign companies, who have invested time, technology, and funds in Russia, by conveniently equating them and aligning them with their home countries, and then completely divesting their assets in the process under the color of law. This tactic, disguised as a legitimate exercise of government power, particularly within its borders, was questioned by some, and was criticized early on. The most recent manifestation of this policy, it appears, has no one fooled.
Rafael X. Zahralddin-Aravena, Partner, Lewis Brisbois Bisgaard & Smith LLP, (“LBBS”), Wilmington, Delaware. Opinions, if any, expressed in this article are his and not the opinion of LBBS. ↑
The most recent amendments instituted a ban against filing bankruptcy against certain classes of debtors to act as a stabilizing feature for the domestic economy during the COVID outbreak. Federal Law No. 98-FZ dated April 1, 2020, “On Amending Certain Legislative Acts of the Russian Federation Regarding Prevention and Liquidation of Emergencies.” Bankruptcy is also governed by the Civil Code of the Russian Federation (the “Russian Civil Code”); the Commercial Procedure Code; and the Law on Enforcement Proceedings. The participation of state authorities in bankruptcy proceedings is also regulated by Resolution No. 257 of the Russian Federation Government dated May 29, 2004, “On Protecting the Interests of the Russian Federation as a Creditor in Bankruptcy Cases and in Bankruptcy Proceedings.” ↑
3 Collier International Business Insolvency Guide P 38.05 (2022). ↑
Sabrina Tavernise, “Using Bankruptcy As a Takeover Tool; Russian Law Puts Healthy Companies at Risk,” New York Times, October 7, 2000, at 1; “Russian Bankruptcy Law Updated,” ABI Journal, September 1999. “Under the Bankruptcy Law, 5,300 bankruptcy cases were filed between March 1, 1998 and December 25, 1998, whereas only 4,200 cases were under the courts’ consideration before March. Furthermore, bankruptcy procedures of the Bankruptcy Law were applied to more than 900 cases opened before March.” ↑
Federal Law of the Russian Federation, No. 6-FZ, “On Insolvency (Bankruptcy),” January 8, 1998. See Sobr. Zakonod., RF, 1998, No. 2, Art. 222. The prior statute was the Law of the Russian Federation “On Insolvency (Bankruptcy) of Enterprises” No. 3929-1, November 19, 1992. See Vedomosti, Verkh, Soveta RF, 1993, No. 1, Art. 6. ↑
See, e.g., Tavernise, at 1; William P. Kratzke, Russia’s Intactable Economic Problems and the Next Steps in Legal Reform: Bankruptcy and the Depoliticization of Business, 21 Nw. J. Int’l L. & Bus. 1, 2 (2000) (“[T]he most important legal reforms for Russia are those that eliminate the reward system that encourages economic activity that can be highly inefficient. These legal reforms are an effective bankruptcy law and the de-politicization of business. The two go hand-in-hand. It is the politicization of business that renders Russia’s bankruptcy laws ineffective by making non-viable business entities appear to be solvent. These two reforms, were they adequately implemented, would eliminate rewards for inefficiency.”). ↑
See, e.g., Kratzke at 30–39; Vassily V. Vitryansky, “Insolvency and Bankruptcy Law Reform in the Russian Federation,” 44 McGill L.J. 409 (August 1999) (Deputy Chairman of the Higher Court of Arbitration of the Russian Federation discussing the necessity for bankruptcy reform and the process of drafting and features of the new insolvency law). ↑
3 Collier International Business Insolvency Guide P 38.05 (2022). ↑
Id. There were also amendments regarding residential property developers and credit institutions (specifically incorporating banks and credit institutions into the Russian Bankruptcy Laws, with certain provisions specific to financial institutions not applicable to other business entities, but a repeal of the prior Federal Law No. 40-FZ on Bankruptcy of Credit Organizations). ↑
Federal Law of the Russian Federation No. 154-FZ, “On the Regulation of the Particulars of Insolvency (Bankruptcy) Within the Territory of the Republic of Crimea and the Federal City of Sevastopol and on the Introduction of Amendments to Certain Legislative Acts of the Russian Federation,” arts. 2–4, 6–14, Ros. Gaz. (July 3, 2015). ↑
Bernard S. Black and Anna S. Tarassova, Institutional Reform in Transition: A Case Study of Russia, 10 S. Ct. Econ. Rev. 211, 253 (“Laws that aren’t honestly enforced are little better, and can sometimes be worse, as those without scruples manipulate the laws for personal gain.”). ↑
Paul B. Stephan, Taxation and Expropriation—The Destruction of the Yukos Oil Empire, 35 Hous. J. Int’l L. 1, 16 (2013) (citing to Allen C. Lynch, Vladimir Putin and Russian Statecraft 58–61 (2011)). ↑
Id. at 2 (citing to various sources which have examined the Yukos saga, the law review literature includes Paul M. Blyschak, Yukos Universal v. Russia: Shell Companies and Treaty Shopping in International Energy Disputes, 10 Rich. J. Global L. & Bus. 179 (2011); Sara C. Carey, What Do the Recent Events Involving Yukos Oil Company Tell Us About Legal Institutions for Transition Economies?, 18 Transnat’l L. 5 (2004); Dmitry Gololobov, The Yukos Tax Case or Ramsay Adventures in Russia, 7 Fla St. U. Bus. Rev. 165 (2008); Dmitry Gololobov, The Yukos Money Laundering Case: A Never-Ending Story, 28 Mich. J. Int’l L. 711 (2007); Matteo M. Winkler, Arbitration Without Privity and Russian Oil: The Yukos Case Before the Houston Court, 27 U. Pa. J. Int’l Econ. L. 115 (2006); Brenden Marino Carbonell, Comment, Cornering the Kremlin: Defending Yukos and TNK-BP from Strategic Expropriation by the Russian State, 12 U. Pa. J. Bus. L. 257 (2009); Peter C. Laidlaw, Comment, Provisional Application of the Energy Charter as Seen in the Yukos Dispute, 52 Santa Clara L. Rev. 655 (2012); Alex M. Niebruegge, Comment, Provisional Application of the Energy Charter Treaty: The Yukos Arbitration and the Future Place of Provisional Application in International Law, 8 Chi. J. Int’l L. 355 (2007)). ↑
See Amoco v. Iran, Award, July 14, 1987, para. 192. (“[A] clear distinction must be made between lawful and unlawful expropriations, since the rules applicable to the compensation to be paid by the expropriating State differ according to the legal characterization of the taking.”). ↑
7 The Chorzów Factory Case (Germany/Poland), September 13, 1928, Series A, No. 17 (substantive issue) at 2. ↑
See Starrett Housing v. Iran, Interlocutory Award No. ITL 32-24-1, December 19, 1983, 4 Iran-United States Claims Tribunal Reports 122, 154 (stating “…it is recognized under international law that measures taken by a State can interfere with property rights to such an extent that these rights are rendered so useless that they must be deemed to have been expropriated, even though the State does not purport to have expropriated them and the legal title to the property formally remains with the original owner.”); The Factory at Chorzów (Claim for Indemnity) (The Merits), Germany v. Poland, Permanent Court of International Justice, Judgment, September 13, 1928, 1928 P.C.I.J. (ser. A) No. 17, at 47; and Norwegian Shipowners’ Claims, Norway v. the United States, Permanent Court of Arbitration, Award, October 13, 1922. ↑
Expropriation UNCTAD Series on Issues in International Investment Agreements II (2012) at 11. ↑
Generation Ukraine v. Ukraine, Award, September 16, 2003, para. 20.22. 6. ↑
Suez et al. v. Argentina, Decision on Liability, July 30, 2010, para. 121. ↑
Expropriation UNCTAD Series on Issues in International Investment Agreements II (2012) at 15. ↑
Id. The Iran-United States Claims Tribunal found expropriation where the Iranian Government appointed temporary managers in the subsidiaries of United States companies and the acts of the government appointees. ↑
“Правкомиссия одобрила проект по национализации имущества ушедших из РФ западных компаний” [“Government commission approved a project to nationalize the property of Western companies that left the Russian Federation”], Russian News Agency TASS, March 9, 2022, https://tass.ru/ekonomika/14012987. ↑
Decree of the President of the Russian Federation dated March 5, 2022, No. 252, “On the application of retaliatory special economic measures in connection with unfriendly actions of certain foreign states and international organizations,” http://actual.pravo.gov.ru/text.html#pnum=0001202205030001. ↑
GOR, Resolution No. 295,431-p of March 6, 2022, “Government approves rules for transactions with foreign companies from unfriendly countries and territories,” March 7, 2022; see also 2022 ITA Unpub LEXIS 343. ↑
Decree of the Government of the Russian Federation dated March 6, 2022, No. 299: “On Amending Paragraph 2 of the Methodology for Determining the Amount of Compensation Paid to the Patentee when Deciding to Use an Invention, Utility Model or Industrial Design without His Consent, and the Procedure for Its Payment,” http://actual.pravo.gov.ru/text.html#pnum=0001202203070005. ↑
On March 30, 2022, Russian Prime Minister Mikhail Mishustin announced and signed a law that made it legal to import grey market goods without proving the legitimacy of the products or obtaining the authorization of the trademark owner (which previously was required). Meeting of the Presidium of the Government Commission to Increase the Sustainability of the Russian Economy under the Sanctions, March 30, 2022, http://government.ru/en/news/44982/. The Russian Ministry of Industry and Trade will be issuing a list of products allowed into Russia, regardless of trademark protections. Id. ↑
See e.g. SwissInfo, “Court rules against Russia over Yukos affair,” August 23, 2007, https://www.swissinfo.ch/eng/court-rules-against-russia-over-yukos-affair/6068860 (noting that the Swiss Federal Court would not allow bank records to be sent to Russia because “Russia’s judicial standards fell short of the international norms needed to comply with the request.”); see also Paul B. Stephan, Taxation and Expropriation—The Destruction of the Yukos Oil Empire, 35 Hous. J. Int’l L. 1, 29–30 (2013) (citing to extradition cases Government of the Russian Federation v. Maruev and Chernysheva, Bow Street Magistrates Court, March 18, 2005; Government of the Russian Federation v. Temerko, Bow Street Magistrates Court, December 23, 2005; Regarding Law on Extraction of Fugitives 95/70, Application No. 2/07, District Court of Nicosia, April 10, 2008; and cases that refused to allow the seizure of corporate records Khodorkovsky v. Office of the Attorney General, Federal Supreme Court, Case 1A29/2007, August 23, 2007; Decision of February 6, 2006, Princely Court of Justice of Principality of Liechtenstein). ↑
Judgment of the District Court of Amsterdam in Case No. 355622/HA ZA 06-3612 of October 31, 2007. On appeal, the Amsterdam Gerechtshof (Court of Appeals) (confirming that Dutch law would not recognize any interest in Yukos Finance that Promneftstroy was granted in the Russian proceeding) and Judgment in Cases No. 200.002.097/01 and 200.002.104/01 of October 19, 2010. ↑
See 11 U.S.C. § 1500, et seq. Enacted in 2005, Chapter 15 governs cross-border bankruptcy and insolvency proceedings and was enacted pursuant to the 1997 UNCITRAL Model Law on Cross-Border Insolvency (“Model Law”) that has been enacted by more than 50 countries. ↑
In re Vneshprombank, Case No. 16-13534 (MG), Bank Foreign Main Proceeding Order, ECF Doc. #24. ↑
In re Vneshprombank, Case No 16-13534 (MKV) (Bankr. S.D.N.Y.). Section 1521(a)(4) allows for discovery by a foreign representative who is recognized by the Court. Section 1521(a)(5) “enables a Foreign Representative to take broad discovery concerning the property and affairs of a [foreign] debtor.” In re Foreign Econ. Indus. Bank Ltd., 607 B.R. 160, 170 (Bankr. S.D.N.Y. 2019) (quoting In re Millennium Glob. Emerging Credit Master Fund Ltd., 471 B.R. 342, 346 (Bankr. S.D.N.Y. 2012)). ↑
In re CJSC Automated Services, Case No. 09-16064 (JMP) (Bankr. S.D.N.Y. Nov. 23, 2009) (recognition granted for claims investigation but foreign representative failed to move forward). ↑
In re Rebgun (Yukos Oil Co.), Case No. 06-10775 (RDD) (Bankr. S.D.N.Y. Feb. 28, 2008) [ECF #145] (Chapter 15 case dismissed after provisional relief was granted but before recognition). ↑
In re Yukos Oil Company, 320 B.R. 130, 136 (Bankr. S.D. Tex. 2004). ↑
United States Department of State, U.S. Relations with Russia, Bilateral Relations Fact Sheet, Bureau of European and Eurasian Affairs, September 3, 2021, https://www.state.gov/u-s-relations-with-russia/. Russia and the United States signed an investment treaty in 1992 that was never enacted. Id. ↑
See, e.g. Agreement Between the Government of the Republic of Singapore and the Government of the Russian Federation on the Promotion and Reciprocal Protection of Investments, September 27, 2010. ↑
Convention on the Recognition and Enforcement of Foreign Arbitral Awards (New York, 1958) Commission On International Trade Law. ↑
Seizures by REPO are authorized under the International Emergency Economic Powers Act (IEEPA). 50 U.S.C. §§1701–1707. However, the IEEPA does not allow the U.S. government to take ownership over the assets as it is not a “vesting” statute, unless the U.S. is “engaged in armed hostilities or has been attacked by a foreign country or foreign nationals,” which is not the current state of the conflict in Ukraine. See 50 U.S.C. §1702(c) (the Patriot Act amended the IEEPA). ↑
General Agreement on Tariffs and Trade, October 30, 1947, 61 Stat. A-11, 55 U.N.T.S. 194 (“GATT 1947”). ↑
Terence P. Stewart and Shahrzad Noorbaloochi, “The WTO Panel Report on Article XXI and its Impact on Section 232 Actions,” Washington International Trade Association, April 11, 2019, https://www.wita.org/atp-research/the-wto-panel-report/. ↑
In re Yukos Oil Co., 321 B. R. 396 (Bankr. S.D. Tex. 2005), and Allen v. Russian Fed’n, 522 F. Supp. 2d 167, 171 (D.C. Cir. 2007). ↑
See Paul B. Stephan, Taxation and Expropriation—The Destruction of the Yukos Oil Empire, 35 Hous. J. Int’l L. 1, 36–45 (2013) (excellent discussion of the various available tribunals and the litigants’ approach in each, with the Strasbourg Court and the arbitration available through bilateral investment treaties being most amenable to hearing cases and finding jurisdiction). ↑
Id. (discussing that the Strasbourg Court held on several points in favor of the Russian taxing authority while arbitration panels disagreed with those decisions). ↑
Yukos Capital S.A.R.L. v. OAO Rosneft, Case No. 200.005.269/01, Decision (April 28, 2009) (Amsterdam Gerechtshof). ↑
Stephan at 41. In summary, “[i]t did not have to determine that Russia violated its own laws, much less any of the specific human rights obligations found in the European Convention. Rather, it had to determine whether Russia’s actions, in their entirety, constituted a compensable expropriation. To do so, it only had to determine that the government’s behavior was inconsistent with ‘routine regulatory powers,’ that is normal tax assessment and enforcement.” See also Quasar de Valores S.I.C.A. v. Russian Fed’n, I.I.C. 557 at P 227 (Arb. Inst. Stockholm Chamber of Commerce 2012). ↑
A new rule (12 CFR Part 1002, the “Rule”) issued in final form by the Consumer Financial Protection Bureau (the “CFPB”) earlier this year, though subject to legal challenges and delay, will impose a host of data collection and reporting obligations on lenders to small businesses. This article provides a high-level overview of key considerations for lenders in light of such enhanced future obligations.
A. Statutory Background:
The Rule, issued on March 30, 2023, implements Section 1071 (“Section 1071”) of the Dodd-Frank Wall Street Reform and Consumer Protection Act. Section 1071, which amended the Equal Credit Opportunity Act, requires lenders who receive applications for small business loans to obtain, maintain, and periodically report to the CFPB certain information about applicants. Further, Section 1071 contemplates that the CFPB will make public the information it receives.
B. The Rule:
Application: The Rule’s reporting requirements apply to any financial institution with at least one hundred “covered originations” in each of the two previous calendar years. “Financial institution” is defined broadly to cover a range of entities engaging in “any financial activity”— i.e., this definition, and by extension, the Rule, could apply to both “traditional” bank lenders and alternative or direct lenders or private credit funds. Covered originations are certain types of credit extensions to “small businesses” (defined as businesses with gross revenues of no more than $5 million in the fiscal year preceding the time of determination).
Requirements: Lenders must report to the CFPB three types of data in connection with each consumer request (importantly, whether written or oral) for a covered credit transaction, including any refinancing of existing debt:
data generated by lenders (e.g., method of credit application and actions taken regarding such application);
data collected from applicants or third parties (e.g., credit purpose, amount requested, and details about the applicants’ business); and
demographic data collected from applicants, including minority-owned business status and information about principal owners.
Additionally, lenders must maintain certain policies and procedures, including:
Procedures reasonably designed to obtain a response to each request for applicant-provided demographic and other data.
Methods to recognize and address “indicia of potential discouragement”—i.e., practices that might cause applicants to decline to provide requested information. Note the CFPB issued a statement that it will “use its enforcement and supervisory authorities to focus on covered lenders’ compliance with” this requirement.
A “firewall” so that, with certain exceptions, persons making credit and other relevant determinations regarding a covered application by a small business do not have access to sensitive information provided by applicants pursuant to the Rule.
Unintentional errors occurring despite maintenance of appropriate procedures will not result in violations of the Rule. There is a presumption that unintentional errors below a numerical threshold are bona fide in nature. Safe harbors for certain types of inaccuracies in reported data also apply.
C. Compliance Considerations:
Smaller lenders will have more time to prepare for compliance than larger lenders. Compliance requirements will be phased in, starting in October 2024 for lenders with at least 2,500 covered originations in the aggregate for both 2022 and 2023. This timing may be affected by litigation for some lenders as described below.
Among other things, lenders must:
Analyze potentially covered credit transactions to determine whether the Rule applies and, if so, collect and provide accurate required data. In addition to developing effective routines to ensure careful and timely reporting, this process involves accurately identifying relevant data for each covered transaction, such as:
the purpose and type of credit;
the application method for such credit;
the reportable amounts (applied for and approved amounts);
pricing information; and
as applicable, reasons for approval or denial of such credit.
Collect and accurately analyze information about applicants, including by identifying relevant demographic and other characteristics.
Maintain procedures to “identify and respond to indicia of potential discouragement” in the credit and lending process.
Develop practices to ensure proper maintenance of the above-mentioned “firewall.” This would involve accurately identifying which persons are (or are not) involved in determining whether to extend credit to a given applicant.
D. Concluding Observations:
The Bureau’s Rule implementing Section 1071 has been the subject of considerable controversy. The rulemaking followed a lawsuit filed by community groups seeking to compel the Bureau to promulgate the Rule, and the process generated voluminous comments, many critical of the new burdens the Rule will impose on lenders. Currently, litigation is pending in federal district court in Texas in which the American Bankers Association and the Texas Bankers Association contend the Rule must be set aside. On July 31, 2023, the court in that case preliminarily enjoined the Bureau from enforcing the Rule against ABA and TBA members pending a decision in CFPB v. Community Financial Services Association, in which the Supreme Court will review the decision of the Fifth Circuit that the CFPB’s funding structure is unconstitutional. Since then, multiple other trade groups representing community banks, credit unions, and others have intervened in the Texas case and requested that implementation of the Rule be enjoined as to their members as well. The Kentucky Bankers Association has also filed a new separate lawsuit together with a group of Kentucky-based lenders similarly contending the Rule must be set aside.
As a practical matter, however, small business lenders would be wise to ensure they can meet future compliance obligations under the Rule. As sketched above, the Rule imposes substantial new administrative burdens, and compliance could prove costly and challenging, especially for community and local institutions. The Rule also seems likely to present challenges for larger institutions, whose existing frameworks for collecting demographic and other consumer information (for example, as required by federal fair lending laws) may not map easily onto the Rule’s distinct requirements. Notably, the Rule is part of a growing trend towards greater regulatory scrutiny over commercial lending, with a particular emphasis on small business transactions—in recent years, state regulators in jurisdictions such as California, Utah, New York, and Virginia have enacted laws mandating broad disclosures by commercial lenders. Considering the Rule’s broad applicability, the numerous new requirements it imposes, and the Bureau’s stated intention to make the Rule’s provisions concerning “discouragement” an enforcement priority, industry participants that engage in small business lending should work with counsel to prepare for compliance.
Your client has finally decided it’s time to acquire an AI-based product/service for its business use and has asked you to review the AI vendor’s standard legal terms relating to the purchase. Where do you begin? This article will highlight some of the key legal issues to consider when acquiring an AI product/service and provide certain risk mitigation strategies that can be employed through contractual means.
Do Your Due Diligence!
Before doing the deep dive into the black and white contract terms, it’s critical to ask your client about the prospective AI vendor. Due diligence is a must in this volatile market, so hopefully your client has done their homework. There are many factors to consider and questions to ask. Is the AI vendor a mature company or a start-up? Has it been the subject of any publicly available complaints, such as regulatory investigations (Canadian or international privacy regulators, the US Federal Trade Commission) or lawsuits?
You also need to know the intended use case, i.e., (i) the nature of the intended AI application; (ii) the industry it will serve; and (iii) how your client will use AI product/service, as these considerations will impact your legal advice. Is the product/service consumer-focused, or is it a business-to-business application that the client will use internally? Has the AI vendor put in place transparency measures to promote openness and explainability in the operation of its products? Will the AI product/service make or affect decisions impacting individuals that are subject to specific laws? What is the origin of the AI product/service? You should understand the scope of its source data—was it captured “in-house” or scraped from “publicly available” sources? You should also confirm the proposed AI contract framework, as the standard vendor terms may reference a number of hyperlinked, ever-changing documents, including an order form, service agreement, separate Terms of Use / Terms of Service, Privacy Policy, additional Legal Terms—all of which should be reviewed.
Consider Bias.
AI systems are far from perfect, as shown by some spectacular (and very public) examples of racist chatbots, financial programs that routinely deny certain minority groups credit/mortgages based on their ethnicity, discriminatory hiring practices, and generative AI programs that hallucinate fictional legal cases, to name a few.
Canadian acquirors of AI products/services should filter and consider their purchases against the requirements of pending Bill C-27, Canada’s proposed Artificial Intelligence and Data Act (the “AIDA”),[1] whose purpose is to expressly regulate certain types of AI systems and ensure that developers and operators of such systems adopt measures to mitigate various risks of harm and avoid biased output.[2] While AIDA will only apply to AI systems that are “high impact” systems (terms are as yet undefined), prospective acquirors should still ask the “hard questions” around the vendor’s bias mitigation practices. Does the AI vendor have an internal AI ethics review board? What kinds of data sets have been used in training the AI product/service? Has the company established measures to identify, assess, and mitigate risks of harm or biased output that could result from a client’s use of the product/service? What steps has the AI vendor taken to ensure the quality and accuracy of its data, to ensure that it is class-balanced and unbiased? Was the source of the AI vendor’s data sufficiently diverse, or was the AI system narrowly focused on a small sample of data that could lead to unforeseen and harmful consequences? Has the AI vendor explicitly tested for bias and discriminatory outcomes? If so, how? Does the company have a plain language description of the AI system that states how it is intended to be used, the types of content that it will generate, and the recommendations, decisions, or predictions that it will make, as well as the strategies to mitigate against bias?
Use Rights / Intellectual Property Considerations / Licensing Concerns.
You should review the draft AI contract to ensure that your client has the necessary rights to use the AI service/product as contemplated, including its affiliates and customers, as applicable. It’s critical to drill down in the prospective AI contract to determine what the vendor says about (i) the ownership of its own intellectual property (AI models, tools), including any licensed third-party content; and (ii) who owns the content/output generated by the AI product/service, as applicable (i.e., the vendor or the client). Since laws are still evolving in this area, all desired client rights must be expressly defined in the AI contract. Many AI systems are built on data sets that have been scraped from other publicly available third-party content, which opens these vendors up to prospective litigation, so a positive affirmation in the vendor contract regarding ownership is essential. Look for language in the AI vendor’s contract to ensure that all rights that make up the AI system have been listed and protected, and that the AI vendor has the right to license the AI technology for its intended uses (any restrictions should be carefully noted).
Privacy/Cybersecurity Issues.
AI systems are rife with privacy concerns. They are myriad, and include (i) ensuring that vendors have the legal authority to process personal information used by the AI product/service, particularly that of minors, in relation to the data sets used to train, validate, and test generative AI models; (ii) individuals’ interactions with generative AI tools; and (iii) the content generated by generative AI tools. Similarly, the AI system should contain mitigation and monitoring measures to ensure personal information generated by generative AI tools is accurate, complete, up-to-date, and free from discriminatory, unlawful, or otherwise unjustifiable effects. Detailed questions should be asked as to whether the AI vendor has put in place sufficient technical and organizational measures to ensure individuals affected by or interacting with these systems have the ability to access their personal information, rectify inaccurate personal information, erase personal information, and refuse to be subject to solely automated decisions with significant effects.
It is therefore critical to understand what the AI vendor says about its own privacy/cybersecurity practices, and whether it has incorporated “privacy/security by design” principles in the development of its AI systems. While AIDA has not yet passed in Canada, existing Canadian privacy laws still require vendors to limit the collection of personal information to only that which is necessary to fulfill the specified task and to ensure that the AI system is not indiscriminately grabbing content solely for the vendor’s benefit. AI vendors should incorporate adequate, reasonable security safeguards to protect against threats and attacks against stored data that seek to reverse engineer the generative AI model or extract personal information originally processed in the datasets used to train the models. The standard AI contact should include detailed language relating to comprehensive privacy protection and mandatory breach notification. Ideally, the vendor will also state in its contract that it adheres to meaningful cybersecurity standards, such as NIST (National Institute of Standards and Technology), which just published its AI Risk Management Framework in January 2023. These requirements and accountability measures must also flow down the vendor’s entire AI supply chain, especially when AI models are built upon one another.
As a start, you should review the AI vendor’s privacy policy, service terms, and terms of use, and subject to your client’s agreement, follow-up questions may be required. You will need to develop a clear picture as to how the vendor will use your client’s content/personal information throughout the life cycle of the AI agreement (including post-termination), and whether/how such personal information will be aggregated/deidentified before use. You should also review the AI vendor’s data retention policies and whether they are acceptable based on your client’s existing third-party obligations/relevant industry.
It is worth noting that starting September 22, 2023, Québec’s Law 25 will grant individuals new transparency and rectification rights related to the use of automated processes to render decisions about individuals (“Automated Decision-Making Systems”) that use the personal information of such individuals. An individual will have the right to: (i) be informed when an enterprise uses their personal information; (ii) request additional information on how the individual’s personal information was used to render a decision, as well as the reasons and principal factors and parameters that led the Automated Decision-Making System to render such decision; (iii) request to have the personal information used to render the decision be corrected, and (iv) submit observations with respect to a decision to a member of the enterprise to review the decision made by an Automated Decision-Making System.
Lastly, it is important to be aware of any “reverse” privacy/security requirements that the AI vendor may incorporate in its standard agreement that create onerous burdens on clients. These may include obligations for clients to notify the vendor of any vulnerabilities or breaches related to the client’s AI service/product and provide details of the breach, provide legally adequate privacy notices, and obtain necessary consents for the processing of client data by the AI vendor, complete with actual representations from the client that they are processing such data in accordance with applicable law. Some AI vendors even require clients to sign separate Data Processing Addenda. It is important to be aware of these additional vendor data requirements and neutralize any that are unacceptable to your client.
Additional Sources of Liability.
Besides the risks above, additional sources of liability include noncompliance with both AI-specific legislation and regulations (which are not limited to Canada, given pending AI regulations in Europe and the United States), but also existing federal and provincial laws (privacy, consumer protection legislation, consumer disclosure requirements). Old laws still continue to apply to AI vendors, and AI systems that are defectively designed would still be subject to product liability laws.
Representations/Warranties/Disclaimers.
Unfortunately, AI products/services are usually offered by vendors on an “as is, as available” basis, with minimal to no legal representations and warranties. Standard contract terms typically contain disclaimers that limit any damages to direct damages with very low dollar liability. You should therefore seek to include express legal representations/warranties regarding the following: (i) the vendor having all necessary rights, including ownership and licenses to make the AI service/product available to the client and for the client to use the AI product/system as contemplated/described; (ii) non-infringement, including no infringement when used by the client as intended; (iii) vendor’s (and the service’s/product’s) compliance with all applicable laws, including privacy laws and jurisdictions outside of Canada (customize as required); (iv) the AI service/product not containing any viruses, malware, etc. that would otherwise damage the client’s systems; and (v) no pending third-party claims or investigations existing that would impact the vendor’s ability to provide the product/service.
Indemnities.
Similarly, many AI vendors do not provide indemnities in their standard legal agreements but rather include reverse indemnities from the client. For example, clients are asked to indemnify the vendor, its affiliates, and personnel from and against claims, losses, and expenses (including legal fees) arising from or relating to: the client’s use of the AI services/product, client’s content, any products or services that the client develops or offers in connection with the AI services or product, or client’s breach of vendor’s terms or applicable law. You should endeavor to minimize the client’s indemnities and balance the agreement through the addition of such critical vendor indemnities as indemnification for vendor’s failure to comply with applicable laws, fraud, negligence/gross negligence, willful misconduct, intellectual property infringement (especially patent and copyright), breaches of confidentiality/privacy and cybersecurity breaches, customer data loss, and lastly, personal injury/death (depending on the product/service). While I do not recommend trying to seek unlimited indemnities as they are generally no longer considered “market,” I recommend instead seeking “super-caps” (i.e., higher caps) for the most critical of these, such as IP infringement; confidentiality breaches and privacy and cybersecurity breaches; customer data loss; fraud; gross negligence/negligence; and willful misconduct. These super-caps may be based on the greater of a specific dollar value or a multiplier based on contract fees paid or payable, or some other formula. Lastly, the scope of the indemnity should include affiliates, contractors, and third-party representatives of the AI vendor as applicable/appropriate.
Dispute Resolution.
You should review what the standard legal agreement says regarding dispute resolution, as many AI vendors seek to restrict a customer’s rights at law (and equity) to deny their day in court. Instead, vendors will insist on mandatory arbitration, naming a US arbitration regime that will prove expensive for the client should it wish to assert its contractual rights. Some agreements also include compelled informal dispute resolution that results in a hold period (i.e., sixty days) before a client can assert a claim. These restrictions may not be in the best interest of the client and should be removed. It is, therefore, important to look at the governing law/jurisdiction clauses carefully and note any special restrictions/differing rights depending on the client’s jurisdiction.
Termination Considerations.
Lastly, don’t forget to look at the termination provisions, as AI contracts often contain robust termination rights in favor of the vendor—i.e., the vendor can terminate the agreement immediately upon notice to client if the client (allegedly) breaches its confidentiality/security requirements, for “changes in relations with third-party technology providers outside of our control,” or to comply with government requests. Also, the vendor may have broad suspension rights that allows suspending the client’s use of the AI system if client is allegedly not in compliance with the AI product/service terms, the client’s use poses a security risk to the AI vendor or any third party, if fraud is suspected, or if the client’s use subjects the AI vendor to liability. Often these broad rights require additional negotiation and tightening to balance the client’s interests. It is also important for the contract to expressly address, in plain language, what happens following contract termination. For example, must the client immediately stop using the service/product and promptly return or destroy the AI vendor’s confidential information? If so, does this include the client’s outputs? Does the client have ongoing usage rights regarding outputs? Will the AI vendor continue to use any ingested client content or personal information, or will this be erased? If yes, consider the protections/restrictions necessary for your client to comply with applicable privacy laws and any particular industry requirements.
Conclusion.
While AI technology may be new, seeking to create balanced legal agreements that correctly apportion risk and liability is not. Notwithstanding the daunting list of risks associated with the use of AI systems, there are a number of risk mitigation measures that prospective buyers (and their counsel) can deploy to manage these concerns. It is critical to negotiate AI contracts with teeth in order to ensure that clients will feel comfortable acquiring and using these products and services on a going-forward basis.
An Act to enact the Consumer Privacy Protection Act and the Personal Information and Data Protection Tribunal Act and to make consequential and related amendments to other Acts, also known as the Digital Charter Implementation Act, 2022 (First Session, Forty-fourth Parliament, 70-71 Elizabeth II, 2021-2022). Bill 27 is comprised of three parts: Part 1 will enact the Consumer Privacy Protection Act; Part 2 will enact the Personal Information and Data Protection Tribunal Act; and Part 3 will enact the Artificial Intelligence and Data Act. ↑
The AI Act defines “biased output” to mean content that is generated, or a decision, recommendation, or prediction that is made, by an artificial intelligence system and that adversely differentiates, directly or indirectly and without justification, in relation to an individual on one or more of the prohibited grounds of discrimination set out in section 3 of the Canadian Human Rights Act, or on a combination of such prohibited grounds. It does not include content, or a decision, recommendation, or prediction, the purpose and effect of which are to prevent disadvantages that are likely to be suffered by, or to eliminate or reduce disadvantages that are suffered by, any group of individuals when those disadvantages would be based on or related to the prohibited grounds. ↑
This article discusses a Showcase CLE program that took place at the ABA Business Law Section’s Fall Meeting on September 7, 2023. All Showcase CLE programs were recorded live and will be available for on-demand credit, free for Business Law Section members.
Over the last several years, issues and trends relevant to the focus on an in-person versus remote workforce and the fate of workers in complex, multitier supply chains have changed dramatically. This session will provide an overview of recent litigation examining corporate directors’ and officers’ (D&O) duties and an overview of emerging legislation addressing workplace misconduct and human rights abuses in supply chains—and it will include a discussion of risk assessment and allocation in the context of mergers and acquisitions (“M&A”).
On September 7, a panel discussion at the American Bar Association Business Law Section Fall Meeting will consider the dramatic direct and indirect changes in the workforce over the past six years, from #MeToo to the pandemic to the “Great Resignation.” The panelists hope to explain the impact of these evolving dynamics on reactions to workplace misconduct and human rights abuses in a variety of different workplaces. The panelists, offering employment law, corporate governance, M&A, and key in-house perspectives, will explain how both the law and best practices have changed in response, and they will provide insights about what business lawyers should advise their clients to do to keep up with current employee/shareholder/stakeholder expectations, conduct appropriate due diligence in these areas, and adhere to new legal requirements.
In the wake of the #MeToo movement, the so-called “#MeToo Representation and Warranty” proliferated in M&A agreements: by 2021 more than half of deals valued over $25 million included specific language focused on sexual harassment or misconduct. The panelists will explain why, despite good intentions, common iterations of the #MeToo representations fail to adequately address the subtle and complex factors that allowed sexual misconduct to proliferate in silence at the Weinstein Company and many other workplaces for decades. The panelists will also clarify how their understanding of effective #MeToo representations has evolved over the past several years, and they will expound on what attorneys can advise their clients to do to ensure that they are adequately addressing liability risk around workplace misconduct in the context of a corporate transaction. The presentation will expand upon what due diligence best practices and emerging trends look like; how thinking has evolved on these efforts since #MeToo went viral in 2018; and what should be done in preparation before the closing of any transaction, including as part of effective integration and policy alignment.
Turning to national and global workforce investigations and high-profile inquiries into corruption, harassment, modern slavery, and other workplace issues related to human rights abuses, the program will expand upon what corporate leaders need to keep in mind as they assess risks associated with a global workforce, especially in the context of transactions. The discussion will include an explanation of the evolution of soft law, like the United Nations Guiding Principles on Business and Human Rights (“UNGPs”) and the Organisation for Economic Co-operation and Development Guidelines for Multinational Enterprises (as recently amended), to hard law, which ranges from “name and shame” frameworks like the United Kingdom’s Modern Slavery Act to mandatory human rights due diligence obligations under France’s Duty of Vigilance, Germany’s Act on Corporate Due Diligence, and the European Union’s long-awaited but imminent Corporate Sustainability Due Diligence Directive.
The panelists will provide guidance on how companies can create policies—like a global code of conduct or business ethics policy—that resonate across borders and effectively address legal risks in various jurisdictions. They will also discuss new risk considerations for directors and officers in the workforce context coming out of the recent McDonald’s decisions, such as board fiduciary obligations to oversee workplace and workforce protections, and officers’ duties to monitor operations within their company-specific silo and collect data for presentation to the board. The panelists will illuminate why it so important, not just from a legal and policy standpoint, but also from a business perspective, to get this right.
The program will feature perspectives from Margaret Egan, Executive Vice President and General Counsel of Hyatt Hotels and Resorts, and Mara Davis, Associate General Counsel, Compliance & Ethics, at Zoom—two panelists who work in-house at large, multinational corporations that are active in the transactional space and operating across the globe. These experts will explain how they are dealing with the dynamics described above on a daily basis and how their respective companies have been able to maintain a cohesive workplace culture despite a growing global workforce, and in the context of hybrid work environments that make personal connection more challenging. They will provide advice on how to set the tone from the top and create a cohesive understanding of corporate culture, especially as new entities are acquired and company reach expands globally. Specifically, they will share their experiences working to create effective community and mentoring/apprenticeship opportunities in a hybrid environment and explain how these efforts dovetail with efforts to address workplace misconduct, such as adequately training investigators even in the context of remote work. Finally, they will draw upon their diverse in-house initiatives to provide examples of how to combat human trafficking, protect staff and supply chain workers from human rights abuses (including sexual harassment), and conduct due diligence in vendor procurement. They will share unique practical suggestions like implementing physical accommodations such as alarm buttons and creating an effective operational-level grievance mechanism as required by the UNGPs.
Together with Egan and Davis, the other panelists—Ally Coll, the founder and CEO of The Purple Method; Harry Jones, a shareholder at Polsinelli; and Charlotte May, a partner at Covington & Burling LLP—will thread these diverse topics together to address how to build a culture of trust and transparency and encourage employees to speak up about workplace issues, even in a remote or hybrid environment. Attendees will leave with an understanding of recent changes in the law; best practices for risk assessment and allocation in the M&A context; and practical changes in workplace investigations, like the use of e-discovery tools and effective reporting channels to ensure that employees and other workers feel safe and protected from retaliation, regardless of where they physically work.
The program will close with a reminder about the upsides of the growing global and hybrid workforce from a company culture, ethics, and business compliance standpoint. The hope is to leave attendees with practical ideas about how to create safe and empowering workplaces, for their own benefit and in order to advise clients on how to do the same.
The use of financial models and projections in fundraising by pre-revenue companies and those in the early and optimistic stages of their business cycle is almost universal. Often, companies disclose assumptions underlying the models and projections. Many include warnings and disclaimers in fine print further advising readers of the uncertainties behind the business and risk in relying too heavily on the models or projections. In these circumstances, the fundraisers feel secure in the disclosures accompanying the models or projections provided to investors.
But what will those models and projections look like two and three years later if the business hits choppy waters, expected business partners change direction, or regulators become less accommodating? Will investors cry foul and draw the attention of the Securities and Exchange Commission (“SEC”), Department of Justice, or other regulators? Recent actions demonstrate heightened enforcement risk to both issuers and their executives in the use of models and projections in fundraising by early- and middle-business-cycle companies. In this article, we offer suggestions on how to mitigate or reduce the risk.
Recent Actions Highlight Danger in the Use of Projections by Early-Stage Issuers
The explosion of initial public offerings by special purpose acquisition companies (“SPACs”) and their privately held target companies via de-SPAC combinations in 2020 and 2021 led the SEC and certain of its officials to raise concerns about the use of projections in fundraising and business combinations. In March 2022, the SEC published proposed rules intended to enhance investor protections, including additional disclosures accompanying projections shared with potential suitors. While the SEC has yet to adopt the proposed rules, recent enforcement activity reflects heightened staff scrutiny of projections and accompanying disclosures used in fundraising that issuers, their executives, and advisors should consider before sharing projections and related disclosures, whether or not connected to business combinations or de-SPAC IPOs.
A client of the authors’ firm, a former executive of a technology company (“TechCo”) that went public via a business combination and de-SPAC transaction, was recently issued a Wells letter notifying him of the SEC staff’s recommendation that the Commission authorize an enforcement action against him. The staff alleges TechCo, and our client, presented misleading financial projections to investors in connection with the business combination and related private investment in public equity (“PIPE”) offering in presentations and filings with the Commission. Specifically, the staff alleges TechCo presented specific revenue projections for the two ensuing calendar years from a particular service unit of the business. Within five months of the first presentation of the projections, however, the targeted business failed to materialize and was later abandoned. Nonetheless, the staff alleges TechCo continued to include the revenue projections for several more months in filings with the Commission.
Ultimately, the staff alleges the projections shared with investors omitted disclosure of assumptions and facts that, in the staff’s view, call into question the reasonableness of the projections. This may sound like a routine SEC enforcement action based on omissions of material facts. Not so fast. A closer look at disclosures and access provided by TechCo to investors reflects a transparency that would ordinarily seem to ward off allegations of any intent to defraud. For example, TechCo’s disclosures with the projections included:
Warning that the projections were based on TechCo management’s “discussions with such counterparties and the latest available information.”
Explanation that the partnerships underlying the projections depended on negotiations over definitive agreements “which have not been completed as of the date of this presentation.”
Confirming once again that descriptions of the business partnerships behind the projections remained “subject to change.”
Arranging for a due diligence call between investors and the potential partner behind the revenue projections at issue.
These disclosures reflect that, at a minimum, investors were told the projections were based on “potential” partnerships, and not signed agreements and committed sales. Moreover, evidence showed that discussions between TechCo and some potential partners continued contemporaneously with most of the filings that included the projections. Nonetheless, the staff alleges that when the partnership negotiations failed to advance as quickly as hoped or expected, continued reference to the projections in filings with the Commission became unreasonable and fraudulent under the federal securities laws.
We often see similar analysis of financial projections and related disclosures in enforcement actions and shareholder disputes where the business fails to develop as expected or as quickly as believed at the time of the share sales. Investors become disillusioned, and recollections of what was disclosed and understood often sour with the passage of time. The recent uptick in enforcement actions regarding technology and early-stage companies, including those that went public via de-SPAC transactions, suggests the clampdown on the use of projections in these circumstances has arrived.
What can an issuer and its executives do to mitigate these risks? Projections are deeply important to investors, and it is unrealistic to propose that companies stop using them, but several strategies could mitigate risk.
Strategies to Mitigate the Risk in Sharing Projections
These strategies to mitigate risk will not close the door to regulatory scrutiny. They can, however, potentially bolster defenses to accusations of deception or misconduct.
1. Title the relevant analysis a “model” and not a “projection.”
If an economic forecast is used in fundraising, title it as a “model” of the business opportunity and not a “projection” of future revenue and income. A “model” describes an analytical tool. A “projection” connotes a prediction or forecast more typical of public companies. Use the “model” description in all communications as well.
2. Do not date the business years in the model.
Date the business progression in the model as Year 1, Year 2, Year 3, etc. Dating the model with specific years (2024, 2025, etc.) can feed into a later characterization of the model as a prediction and provide target dates that can be used against you.
3. Define and disclose when Year 1 in the model begins.
This is particularly important for pre-revenue and early business cycle companies. Does Year 1 begin with the public launch of a new product, adoption of the product by certain partners, regulatory approval of a new product, or clearance of some other hurdle or obstacle to the launch of the business? Defining the start of Year 1 can help guard against unexpected delays in developing the business.
4. Disclose assumptions behind the model.
Disclosing assumptions built into the model can guard against future efforts to allege management concealed factors that years later may appear unreasonable in light of subsequent micro- or macroeconomic events. This should go beyond disclosure of just the Excel spreadsheet or other raw data behind the model. If possible, define terms and provide narrative explanations of the reasoning and bases behind the model. This can help to guard against claims of confusion about the meaning of the data and calculations in the raw model spreadsheet.
5. Document all meetings and presentations relating to the model.
This seems obvious, but we have seen numerous failures by management to memorialize oral representations to potential investors that accompany the presentation of projections and models. Noting questions raised by potential investors can also aid later efforts to reconstruct what was important to investors at the time. This can take the form of a formal transcript or notes from the meeting or presentation.
6. Consider a risk review before sharing the model with investors.
If resources permit, it may be beneficial to have experienced SEC enforcement counsel review the business model and related financial due diligence before disclosure to investors. In addition to identifying potential areas of weakness or exposure, a review may provide a potential advice-of-counsel defense to the company and executives, so long as all related facts are disclosed to reviewing counsel.
While these strategies seem simple, we continue to see enforcement actions aimed at the use of financial projections where some or none of the above strategies were implemented. The SEC’s announced intent to more closely police the use of projections in fundraising, as well as the aggressive positions taken by the staff in the case of TechCo and our client, counsel for greater caution.
This article is related to a Showcase CLE program that took place at the American Bar Association Business Law Section’s Fall Meeting on September 8, 2023. All Showcase CLE programs were recorded live and will be available for on-demand credit, free for Business Law Section members.
In March 2023 the United States experienced two of the largest bank failures in its history, Silicon Valley Bank (“SVB”) and Signature Bank, with the failure of First Republic Bank following shortly thereafter.[1] This article reviews aspects of these failures (mostly of SVB’s failure)—in particular, the effect of rising interest rates, including on longer-duration securities; some of the precipitating events; and whether the failures provide lessons about the regulatory tools that might have led to a different outcome. This article seeks to expand the dialogue away from a binary debate centered around questions of whether the 2023 turmoil means that policymakers should raise (or not) capital requirements and unwind (or not) tailoring of the prudential framework that was undertaken in 2018–2019—largely in response to Congress passing the Economic Growth, Regulatory Relief, and Consumer Protection Act of 2018.[2] In an era when questions about the politicization of the federal banking agencies abound, it is worth asking whether there are some policy approaches that can be pursued that avoid, and indeed are more targeted than, some of the policy questions that have become the center of political debates.[3] If achieving that perhaps aspirational goal is possible, it should allow for durable policy and redound to the benefit of finding the right balance between financial stability and economic growth and innovation.
One way to undertake this exercise is by taking a fresh look at aspects of the Dodd-Frank Act (“DFA”) that were never implemented (particularly, section 166); reviewing what the banking agencies previously said about how to capitalize unrealized losses on investment securities; and examining particular attributes of SVB’s failure. These points then can be considered through the lens of what regulators have said about how to respond to the March 2023 banking sector turmoil. Evaluations of this nature are useful because they can help inform how to design regulatory and policy responses that are targeted to address the particular lessons learned from these events.
Accordingly, this article reviews a series of approaches that could be considered to address the attributes of the March 2023 banking sector turmoil and would be less drastic and disruptive than wholesale changes to the prudential regulatory framework, including the regulatory capital standards that apply to large banking organizations. These approaches are revisiting early remediation requirements; revising how unrealized losses on investment securities are capitalized; and designing new triggers to better prepare for the resolution of large banking organizations and, in turn, to develop a more effective resolution paradigm.
By no means are these approaches intended to be presented as the exclusive or best policy approaches that may be pursued in response to the March 2023 banking sector turmoil. Instead, this article seeks to illustrate two main points: one, it is hard to say that the March 2023 turmoil demonstrates the DFA was structurally flawed, given that the agencies have not implemented key provisions of that law; and, two, it is worth considering whether there are targeted responses that would be able to address the problems that were revealed with relative efficiency.
Looking Back at Proposed, but Unadopted, Regulatory Measures
2010: Dodd-Frank Act
In response to the 2008 financial crisis, Congress passed the Dodd-Frank Act “[t]o promote the financial stability of the United States.”[4] This section discusses two of the DFA’s directives to help frame the remainder of the article: first, a study commissioned to understand the effectiveness of prompt corrective actions (“PCA”); and second, a mandate for regulations providing for the early remediation of large, interconnected financial companies facing financial distress.[5]
The PCA regime was adopted in 1992 and “implement[ed] a statutory requirement that banking regulators take specified ‘prompt corrective action’ when an insured institution’s capital falls to certain levels.”[6] As was evidenced by the crisis in 2008, however, there were fundamental weaknesses in the tools used by regulators to deal promptly with emerging issues at the time.[7] These observed weaknesses prompted a study commissioned under the DFA that, among other findings, “recommend[ed] that the bank regulators consider additional triggers that would require early and forceful regulatory action to address unsafe banking practices as well as the other options identified in the report to improve PCA.”[8] Furthermore, section 166 of the DFA was designed to address these same types of concerns.
In particular, section 166 of the DFA requires that, among other things, the Federal Reserve Board (“FRB”) prescribe regulations establishing standards for the early remediation of large, interconnected bank holding companies under financial distress.[9] In 2012, the FRB proposed a rule to implement this provision. At the time, the FRB said:
The recent financial crisis revealed that the condition of large banking organizations can deteriorate rapidly even during periods when their reported capital ratios are well above minimum requirements. The crisis also revealed fundamental weaknesses in the U.S. regulatory community’s tools to deal promptly with emerging issues. As detailed in the Government Accountability Office’s (GAO) June 2011 study on the effectiveness of the prompt corrective action (PCA) regime, the PCA regime’s triggers, based primarily on regulatory capital ratios, limited its ability to promptly address problems at insured depository intuitions. The study also concluded that the PCA regime failed to prevent widespread losses to the deposit insurance fund, and that while supervisors had the discretion to act more quickly, they did not consistently do so. Section 166 of the Dodd-Frank Act was designed to address these problems by directing the Board to promulgate regulations providing for the early remediation of financial weaknesses at covered companies.[10]
The FRB’s proposal would have required a series of early remediation triggers and requirements cascading in stringency, from level one through level four.[11] Level one, or heightened supervisory review, would have been triggered when a firm first shows signs of financial distress such that the firm is likely to experience further decline.[12] Level two, or initial remediation, would have imposed limits on capital distributions, acquisitions, and asset growth for those banks.[13] Of note, at level two, a firm’s assets would have been limited to growing by no more than 5 percent quarter-over-quarter and year-over-year. Level three, or recovery-level remediation, would have required, among other things, development of a capital restoration plan; broad limits on the ability to conduct business as usual; and, importantly, a written agreement with the FRB prohibiting capital distributions, asset growth, and material acquisitions. Furthermore, for level three, a firm would have been subject to a prohibition on discretionary bonus payments and restrictions on pay increases, and supervisors would have had the ability to remove culpable senior management and limit transactions between affiliates.[14] At level four, the FRB would have considered whether to recommend resolution for the firm.[15]
Although these measures ultimately went unadopted, when the FRB implemented the DFA’s enhanced prudential standards in 2014, the FRB said that early remediation requirements would be adopted at a later date following further study.[16] It is not clear whether such a study occurred and, if so, what it concluded.
2013: Basel III Final Rule
When the federal banking agencies began to implement the Basel III capital standards in 2012, they proposed that all banking organizations be required to include certain aspects of accumulated other comprehensive income (“AOCI”) in regulatory capital.[17] AOCI “generally includes accumulated unrealized gains and losses on certain assets and liabilities that have not been included in net income, yet are included in equity under U.S. generally accepted accounting principles (GAAP) (for example, unrealized gains and losses on securities designated as available-for-sale (AFS)).”[18] AOCI is recorded in the equity section of the balance sheet, and, therefore, unrealized losses recorded in AOCI reduce equity.[19] The agencies believed that this proposed AOCI treatment would result in “a regulatory capital measure that better reflects banking organizations’ actual loss absorption capacity at a specific point in time.”[20] In response to the proposed rule, however, the agencies received comments asserting that the proposed treatment would result in “volatility in regulatory capital” and “significant difficulties in capital planning and asset-liability management.”[21] Ultimately, the final rule allowed certain banks not subject to the Advanced Approaches capital standards (ultimately, SVB was among them) to opt out of having AOCI flow through to regulatory capital.[22]
In making this decision, the agencies noted that
while the agencies believe that the proposed AOCI treatment results in a regulatory capital measure that better reflects banking organizations’ actual loss absorption capacity at a specific point in time, the agencies recognize that for many banking organizations, the volatility in regulatory capital that could result from the proposals could lead to significant difficulties in capital planning and asset liability management. The agencies also recognize that the tools used by larger, more complex banking organizations for managing interest rate risk are not necessarily readily available for all banking organizations.[23]
2022: Financial Stability Oversight Council Annual Report
More recently, the Financial Stability Oversight Council (“FSOC”) highlighted these same risks that the agencies observed in 2013. Specifically, the FSOC cautioned in its 2022 annual report that “[i]nvestment portfolios are at risk as [interest] rates rise, . . . [and] a rapid increase in rates may decrease profitability for banks with larger shares of long duration holdings. . . .”[24]
Two charts in the report, reproduced below, highlight the FSOC’s observations. First, the FSOC illustrated that AOCI at the end of 2021 and into 2022 represented negative 15 percent of equity for large complex bank holding companies, above negative 10 percent of equity for large noncomplex bank holding companies, and less than negative 10 percent of equity for U.S. global systemically important bank holding companies (“U.S. GSIBs”).[25] These levels of negative equity sharply increased from 2020 and early 2021. In addition, the FSOC illustrated, correspondingly, that into 2022, U.S. GSIBs held over 60 percent of investment securities as held-to-maturity (“HTM”), whereas large complex and large noncomplex bank holding companies held less than 30 percent and less than 20 percent, respectively, of investment securities as HTM.[26] This illustration is corresponding because HTM securities do not flow through to AOCI; therefore, as a firm holds more HTM securities in its investment portfolio, the investment portfolio will contribute less volatility to AOCI.
Chart A
This chart shows the extent to which unrealized losses on AFS securities resulted in negative equity balances across the banking industry in 2022. Source: FSOC 2022 Annual Report. See footnote 24.
Chart B
This chart shows the percentage of investment securities classified as AFS and HTM and the stark differences in those classifications as between GSIBs, on the one hand, and large complex and large noncomplex bank holding companies, on the other. Source: FSOC 2022 Annual Report. See footnote 24.
Silicon Valley Bank
As of December 2022, just three months before SVB’s failure, SVB Financial Group (“SVBFG”), SVB’s parent holding company, had total assets of just over $200 billion and had invested nearly half those assets in HTM securities like Treasury bonds.[27] As the FRB raised interest rates, SVBFG experienced a dramatic loss in the value of its investment security portfolio; however, as noted above, under regulatory capital standards in effect at the time, this value leakage did not affect SVBFG’s regulatory capital ratios.[28] Indeed, if unrealized losses on SVBFG’s AFS and HTM portfolios had been subtracted from total balance sheet equity and total regulatory capital for accounting and regulatory capital purposes, SVBFG’s balance sheet equity and total regulatory capital would have reflected approximately negative $2.9 and negative $0.65 billion, respectively, in September 2022.[29] The chart below reflects this point.
Chart C
This chart, created by the author, shows the affect investment security losses would have had on SVBFG’s total regulatory capital and total balance sheet equity. See footnote 29.
Further, as stated in FRB Vice Chair Michael Barr’s report on the supervision and regulation of the bank:
On March 9, SVB lost over $40 billion in deposits, and SVBFG management expected to lose over $100 billion more on March 10. This deposit outflow was remarkable in terms of scale and scope and represented roughly 85 percent of the bank’s deposit base. By comparison, estimates suggest that the failure of Wachovia in 2008 included about $10 billion in outflows over 8 days, while the failure of Washington Mutual in 2008 included $19 billion over 16 days. In response to these actual and expected deposit outflows, SVB failed on March 10, 2023, which in turn led to the later bankruptcy of SVBFG.[30]
The deposit outflow on March 9, however, followed a longer, slower-motion outflow of deposits from SVB. Specifically, as reflected in the chart below, from March 31, 2022, until December 31, 2022, SVB’s average total deposits declined from approximately $191 billion to $175 billion, and SVB’s average noninterest-bearing deposits declined from $125.5 billion to $86.9 billion dollars—representing an approximately $17 billion, or 8 percent, and $38.6 billion, or 30 percent, outflow, respectively.[31] That is, the deposit outflow, particularly with respect to uninsured deposits observed in March 2023, in effect began one year prior. Said differently, it also may be possible to frame what occurred as a deposit outflow of approximately $56 billion and $80 billion dollars, respectively, over an approximately twelve-month period, rather than a $40 billion outflow in a single day.
Chart D
This chart, created by the author, shows the deposit outflows experienced by SVB over the year prior to the bank’s failure. See footnote 31.
According to congressional testimony from Federal Deposit Insurance Corporation (“FDIC”) Chairman Martin J. Gruenberg, the FDIC began developing a resolution strategy on the evening of March 9, 2023—just hours before the bank failed.[32] As has been documented rather extensively heretofore, in resolving SVB, the FDIC ultimately invoked the statutory systemic risk exception (“SRE”), which effectively allowed the FDIC to guarantee repayment of all of SVB’s deposits, whether insured or not.[33] The ability to invoke the SRE requires approval by two-thirds of both the FDIC board and the FRB, approval by the Treasury secretary, and consultation with the president.[34] Having successfully invoked this measure, on March 13, the Monday following SVB’s failure, the FDIC stated that “[d]epositors will have full access to their money beginning this morning [and] all depositors of the institution will be made whole . . . both insured and uninsured. . . .”[35] The FDIC has estimated that SVB’s failure will cost the deposit insurance fund $20 billion but noted that “[t]he exact cost will be determined when the FDIC terminates the receivership.”[36]
Select Commentary
The commentary from regulators in the wake of the March 2023 turmoil has included a diverse collection of thoughts, empirical reporting, and proposed solutions for a path forward. Some, including Barr, have focused on the need for stronger capital requirements, saying, “[B]anks with inadequate levels of capital are vulnerable, and that vulnerability can cause contagion.”[37] Others, like FDIC Director Jonathan McKernan, have focused on the idea that “an effective resolution framework [is part of] our best hope for eventually ending our country’s bailout culture that privatizes gains while socializing losses.”[38] And while the Treasury Department’s Assistant Secretary for Financial Institutions Graham Steele approves of the current focus “on the unrealized losses in banks’ available-for-sale and held-to-maturity securities as important metrics to assess a bank’s solvency,”[39] FRB Governor Michelle Bowman has observed that the recent failures rest squarely on “poor risk management and deficient supervision, not . . . a lack of capital.”[40]
Dan Tarullo, a professor at Harvard Law School and a former FRB governor, framed it this way:
I think the agencies need to be especially careful here not to overreact to the events of this spring. It’s of course critical to address the vulnerabilities that were exposed and, as I said earlier, to make sure banks that are undershooting their profit targets do not take excessive risks. But the agencies need to think through whether some ideas for increased regulation would just exacerbate the competitive problems of these banks while not efficiently containing those vulnerabilities.[41]
In the spirit of Tarullo’s comments, the discussion below reviews potential policy tools that are available to address what the spring 2023 turmoil revealed, apart from the broader and more divisive debate about capital calibration and the appropriateness of tailoring the prudential regulatory framework based on the size of a banking organization.
Policy Considerations
Revising Early Remediation
On consideration is whether the FRB should promptly implement the DFA’s early remediation requirements.
Certainly, it should be possible to look back and evaluate how such requirements could have helped avoid the use of the SRE and the hectic resolution of SVB. For example, what triggers could have required swift action as SVB experienced a slow-motion run on 8 percent and 30 percent of its average total and average noninterest-bearing deposits, respectively? If those triggers had been in place, would the March 9 run have been avoided—or at least been less of a surprise? Could triggers be designed that would have prevented the accumulation of SVBFG’s negative equity balance, described above? Or required prompt action once it had accumulated?
Of course, picking the right triggers and remedial actions is no easy task, and it is also worthwhile to avoid fighting the last battle when designing policy. Moreover, remedial actions should be designed to avoid exacerbating a firm’s deteriorating financial condition. Nevertheless, the problems the FRB described when proposing early remediation rules in 2012 (“that the condition of large banking organizations can deteriorate rapidly even during periods when their reported capital ratios are well above minimum requirements”)[42] appear to still be present and to have been a part of the reason why the SRE needed to be used in resolving SVB. Moreover, it is not clear, for example, that given the problem that section 166 is designed to address, whether higher capital requirements would avoid a similar situation in the future.
Accordingly, DFA section 166 seems like a targeted tool that can be evaluated and used to fill the regulatory gaps that March 2023 revealed. In all events, the fact that designing rules involves complicated policy judgments, such as those described above, does not seem like a reason for the agencies to avoid faithfully implementing the laws on the books.[43]
Capitalizing Unrealized Losses on Investment Securities
As also reviewed above, the agencies previously said that having AOCI flow through to regulatory capital results in a capital measure that better reflects banking organizations’ actual loss absorption capacity at a specific point in time. This prior statement appears, in hindsight, to have been correct and worth revisiting, as the agencies recently have proposed.[44] Indeed, in this regard, the way unrealized losses were treated for SVB appears to show that, at least in this respect, capital did in fact play a role in its failure, given that the negative equity position likely caused depositors to have concern about the bank’s financial condition.
That said, the agencies naturally will have to grapple with the question that they previously noted, in particular, whether “tools used by larger, more complex banking organizations for managing interest rate risk are . . . readily available for all banking organizations” and, if not, the size threshold at which having AOCI flow through to capital is not necessary.[45] Making this judgment should involve considering the size above which resolution of an institution is likely to threaten financial stability. This question, however, should not be viewed in isolation. For example, if clear and strong early remediation requirements are in place, as discussed above, and the way in which the FDIC plans for resolution, as discussed below, is enhanced, then the likelihood that the failure of even a relatively large firm would threaten financial stability should be (perhaps materially) lower.
Preparing for Resolution and Developing an Effective Resolution Framework
Another lesson from SVB’s failure is perhaps one of the more obvious—beginning the process to resolve a $200 billion bank the evening before its failure does not provide sufficient runway to conduct an orderly resolution. Thus, the natural question that follows is this: When should the FDIC begin to actively plan for resolution?
For example, if there had been triggers for the FDIC to begin to prepare—such as SVB’s deposit outflows or the dramatically large unrealized losses on SVBFG’s balance sheet and the effective result that had on equity levels—would the SRE have been needed? If the FDIC had begun to prepare for SVB’s resolution, for example, in September 2022 (when the equity balance was negative ~$2.9 billion (see Chart C)) or after year-end 2022 (with average total and average noninterest-bearing deposits down ~8 percent and ~30 percent, respectively, in twelve months (see Chart D)), would the firm’s failure have been easier to manage? Perhaps the FDIC and other regulators would have been able to use this time to identify impediments to a sale, remedy them, and be ready to sell the firm to one or more buyers over a weekend. In addition, this time could have allowed the FDIC and other regulators to evaluate the type of resolution that was best suited to the circumstances. For example, was a resolution of the bank and bankruptcy of the holding company most appropriate, or would invocation of the DFA’s Title II orderly liquidation authority have been useful?
Further, would SVB’s management and board have been spurred to act more swiftly to address the firm’s deteriorating condition if they were advised by the FDIC that the agency was beginning plans for the resolution and sale of the firm? Experience suggests that hearing that message from the FDIC is sobering for management and a board and stiffens the spine to take difficult, and perhaps previously hard to imagine, actions.
Another adjacent question is whether clear and strong early remediation requirements could have worked in tandem with earlier resolution preparedness. Of course, important questions would need to be considered, such as: What are the appropriate triggers for resolution preparedness? Which agency should be responsible for calling in the FDIC to begin that preparation?
Conclusion
All of the above policy considerations, and the others being considered by policymakers, are complex, and different solutions have associated pros and cons. Financial regulatory policy is sufficiently complex that no one proposal is ever likely to provide a magic bullet. The above, however, shows that the DFA had provisions designed to address situations like the one that transpired earlier in 2023, but those provisions were never implemented. To that end, this article aims to put forward for consideration targeted proposals for using those tools in a way that would address the vulnerabilities revealed in the spring of 2023 and would help forge a more resilient financial system—and, in doing so, hopefully avoid, as Tarullo said, an overreaction that could exacerbate broader structural problems facing the banking sector.
This article represents the views of the author, not those of his firm or any client of the firm. The author gratefully acknowledges the assistance of Jeremy R. Lee, associate at Davis Polk, in the preparation of this article. The author also would like to acknowledge with gratitude the willingness of Alex LePore to take the time to challenge and help refine the author’s policy thoughts, whether we agree or disagree, including those thoughts presented in this article.
The Economic Growth, Regulatory Relief, and Consumer Protection Act, S. 2155, 115th Cong. (2018). ↑
See David Wessel, Talking to Dan Tarullo About Bank Mergers, Stress Tests, and Supervision, Brookings (Aug. 10, 2023) (“That’s changed, as so much in the country has changed. Issues that were formerly a little bit blurred have become increasingly partisan. It’s almost as if, when people from one party have a position, there’s a reflexive instinct on people from the other party to oppose that position.”). ↑
Dodd-Frank Wall Street Reform and Consumer Protection Act, H.R. 4173, 111th Cong. (2010) [hereinafter DFA]. ↑
See Enhanced Prudential Standards and Early Remediation Requirements for Covered Companies, 77 Fed. Reg. 634 (proposed Jan. 5, 2012) (to be codified at 12 C.F.R. pt. 252) (“The crisis also revealed fundamental weaknesses in the U.S. regulatory community’s tools to deal promptly with emerging issues.”) [hereinafter Early Remediation Reqs.]. ↑
Recent Bank Failures and the Federal Regulatory Response: Hearing Before the S. Comm. on Banking, Hous., & Urb. Affs., 118th Cong. 7 (Mar. 28, 2023) (statement of Martin J. Gruenberg, Chairman, Federal Deposit Insurance Corporation). ↑
To this end, another of DFA’s requirements that merits revisiting is the unfulfilled obligation of the federal banking agencies to adopt rules regarding the requirement for a bank holding company or savings and loan holding company to act as a source of financial strength for any subsidiary depository institutions. See 12 U.S.C. § 1831o-1. ↑
