Publications

Much has been written recently about artificial intelligence (AI) and law. But what is AI, and what is its relation to the practice and administration of law? This article addresses those questions by providing a high-level overview of AI and its use within law. The discussion aims to be nuanced but also understandable to those without a technical background. To that end, I first discuss AI generally. I then turn to AI and how it is being used by lawyers in the practice of law, people and companies who are governed by the law, and government officials who administer the law. A key motivation in writing this article is to provide a realistic, demystified view of AI that is rooted in the actual capabilities of the technology. This is meant to contrast with discussions about AI and law that are decidedly futurist in nature.

Machine-learning algorithms are improving and automating important functions in medicine, transportation, and business. Government officials have also started to take notice of the accuracy and speed that such algorithms provide, increasingly relying on them to aid with consequential public-sector functions, including tax administration, regulatory oversight, and benefits administration. Despite machine-learning algorithms’ superior predictive power over conventional analytic tools, algorithmic forecasts are difficult to understand and explain. Machine learning’s “black-box” nature has thus raised concern: Can algorithmic governance be squared with legal principles of governmental transparency? We analyze this question and conclude that machine-learning algorithms’ relative inscrutability does not pose a legal barrier to their responsible use by governmental authorities. We distinguish between principles of “fishbowl transparency” and “reasoned transparency,” explaining how both are implicated by algorithmic governance but also showing that neither conception compels anything close to total transparency. Although machine learning’s black-box features distinctively implicate notions of reasoned transparency, legal demands for reason-giving can be satisfied by explaining an algorithm’s purpose, design, and basic functioning. Furthermore, new technical advances will only make machine-learning algorithms increasingly more explainable. Algorithmic governance can meet both legal and public demands for transparency while still enhancing accuracy, efficiency, and even potentially legitimacy in government.

Artificial General Intelligence (“AGI”) - an Artificial Intelligence ("AI") capable of achieving any goal - is the greatest existential threat humanity faces. Indeed, the questions surrounding the regulation of AGI are the most important the millennial generation will answer. The capabilities of current AI systems are evolving at accelerating rates. Yet, legislators and scholars have yet to address or identify critical issues relating to AI regulation. Instead, legislators and scholars have focused narrowly on short term AI policy. This paper takes a contrarian approach to analyzing AI regulation with a specific emphasis on deep reinforcement learning systems, a relatively recent breakthrough in AI technology. Additionally, this paper identifies three important regulatory issues legislators and scholars need to address in the context of AI development. AI and legal scholars have made the demanding need for an AI regulatory system clear. However, those arguments focus on the regulation of current AI systems and generally ignore or dismiss the possibility of AGI. Further, previous scholarship has yet to grapple specifically with the regulation of deep reinforcement learning systems, which many AI scholars argue provides a direct path to AGI. Ultimately, legislators must consider and address the perils and promises of AGI when developing and evolving AI regulatory frameworks.

For the past decades, scholars from varied scientific fields have insisted on the need to develop ‘future-proof solutions’. Future proofing science, design, architecture, and technology entails the creation of solutions that are forward-looking, sustainable, resilient, and can adapt to complex challenges. In the last years, lawmakers have also become intrigued by the question whether laws and policies could also be future-proof. This question has become particularly relevant considering the rapid changes that characterize the digital age. The implementation of a future-proof approach to legislation has nonetheless remained overlooked in the legal literature. This paper aims to fill this gap with an interdisciplinary analysis of future proofing and a discussion of the challenges of implementing this approach to lawmaking. We contend that future proofing law is a challenging task for legislators but a cautious forward-looking approach could ensure that legislation becomes more adaptable and flexible to innovation. Drawing on the interdisciplinary literature on future proofing, we suggest the broader employment of experimental legislation and regulatory impact assessments.

Over the past few years, there has been much discussion regarding the potential of automated-decision making (‘ADM’) systems powered by mechanisms of computational intelligence such as machine learning or deep learning (commonly referred to as ‘Artificial Intelligence’ or ‘AI’). To date, such forms of (big) data analysis are most prominently relied on by the private sector, such as the search algorithms used by online search engines or the recommendation algorithms used by e-commerce and entertainment services platforms. These forms of data analysis in essence offer three main benefits, namely the speed and efficiency of decision-making as well as an ability to detect correlations that may be undetectable to the human brain. The efficiency, speed and correlations offered by these forms of data analytics are also appealing in the public sector. Indeed, various products of computational learning are already being used in administrative processes and will likely become much more prominent in future years. Whereas these techniques offer important potential benefits, they have also been the cause of concern. Indeed, the use of ADM in administrative settings raises numerous important legal and ethical challenges. This paper introduces these new elements in the administrative toolbox and to survey related consequences, in particular possible implications for the principle of transparency.

Cost-benefit analysis (“CBA”) has faced significant opposition during most of its tenure as an influential agency decisionmaking tool. As advancements have been made in CBA practice, especially in more complete monetization of relevant effects, CBA has been gaining acceptance as an essential part of reasoned agency decisionmaking. When carefully conducted, CBA promotes transparency and accountability, efficient and predictable policies, and targeted retrospective review. This Article highlights an underappreciated additional effect of extensive use of CBA to support agency rulemaking: reasonable regulatory stability. In particular, a regulation based on a wellsupported CBA is more difficult to modify for at least two reasons. The first reason relates to judicial review. Courts take a “hard look” at agency findings of fact, which are summarized in a CBA, and they require justifications when an agency changes course in ways that contradict its previous factfinding. A prior CBA provides a powerful reference point; any updated CBA supporting a new course of action will naturally be compared against the prior CBA, and the agency will need to explain any changes in CBA inputs, assumptions, and methodology. The second reason relates to the nature of CBA. By focusing on the incremental costs and benefits of a proposed change, CBA can make it difficult for an agency to justify changing course, especially when stakeholders have already relied on the prior policy. Together, these forces constrain the range of changes that agencies could rationally support. CBA thus promotes regulatory stability around transparent and increasingly efficient policies. But, admittedly, this CBA-based stabilizing influence gives rise to several objections. This Article responds to, among others, concerns about democratic accountability and, most importantly, the use of alternative methods of policy modification. Overall, the Article concludes that CBA and judicial review of CBA play a desirable role in stabilizing regulatory policy across presidential administrations.

Complaints about excessive economic burdens associated with regulation abound in contemporary political and legal rhetoric. In recent years, perhaps nowhere have these complaints been heard as loudly as in the context of regulations targeting the use of coal as an energy source, as production levels in the coal industry dropped nearly by half between 2008 and 2016. The coal industry and its political supporters, including the President of the United States, have argued that a suite of air pollution regulations imposed by the U.S. Environmental Protection Agency (EPA) during the Obama Administration seriously undermined coal companies’ bottom lines, presenting an existential threat to the industry. Under the Trump Administration, industry players have lobbied hard for (and sometimes received) financial subsidies and regulatory changes, with the President seemingly all-too-happy to play the role of the industry’s savior. Stepping back, we ask whether regulations have really led to the decline in demand for coal and how much the coal industry can actually expect to gain from the de-regulatory policies of the current Administration. To address these questions, we statistically analyze stock market reactions to important events in what critics called the regulatory “war on coal” during the Obama Administration. Using an event-study framework that measures abnormal market activity in the immediate wake of these events, we are able to isolate any potential impact of regulation above and beyond market factors, such as secular trends in natural gas prices and market performance as a whole. Surprisingly, we find no systemic evidence consistent with a regulatory “war on coal” based on investor assessments of the industry’s financial prospects, even though our methods do find evidence of stock market reactions to other events, such as bankruptcies of other companies. The very actors with financial stakes in understanding the impact of regulation on the coal industry never bought into the regulatory “war on coal” narrative. Our findings are consistent with other evidence about the effects of regulation and with an underlying political economy of regulatory scapegoating, according to which actors in a declining industry prefer to blame regulation rather than competitive factors for the decline. By recognizing the pervasive incentives for scapegoating and cheap talk by politicians seeking to be saviors, we explain the mismatch between the evidence and the rhetoric of the “war on coal,” and along the way we also show how important it is for courts, government officials, and the public to demand careful analysis and evidence before agencies make regulatory decisions.

The European Commission has recognised the importance of a more innovation-oriented EU acquis, gradually exploring the ways in which EU rules can support innovation. The ‘innovation principle’ was introduced to ensure that whenever policy is developed, the impact on innovation is fully assessed. However, as further discussed in this Study, the exact contours of the innovation principle have been shaped very gradually within the context of the EU better regulation agenda: originally advocated by industry in the context of the precautionary principle, the innovation principle has gradually been given a more articulate and consistent role, which aims at complementing the precautionary principle by increasing the salience of impacts on innovation during all phases of the policy cycle. This Study presents an evaluation of the current implementation of the innovation principle, limited to two of its three components, i.e. the Research and Innovation Tool included in the Better Regulation Toolbox, and the innovation deals. As a preliminary caveat, it is important to recall that the implementation of the innovation principle is still in its infancy, and thus the Study only represents a very early assessment of the extent to which the innovation principle is being correctly implemented, and whether changes would be required to make the principle more effective and useful in the context of the EU better regulation agenda. The main finding is that the innovation principle has the potential to contribute to the quality and future-proof nature of EU policy, but that significant changes and effort will be needed for this potential to fully materialise. The most evident areas for improvement are related to the lack of a clear legal basis, the lack of a widely acknowledged definition, the lack of awareness among EU officials and stakeholders, and the lack of adequate skills among those that are called to implement the innovation principle. As a result of these problems, the impact of the innovation principle on the innovation-friendliness of the EU acquis has been limited so far. The Commission should clarify in official documents that the Innovation principle does not entail a de-regulatory approach, and is not incompatible with the precautionary principle: this would also help to have the principle fully recognised and endorsed by all EU institutions, as well as by civil society, often concerned with the possible anti-regulatory narrative around the innovation principle in stakeholder discussions. Apart from clarifications, and further dissemination and training, major improvements are possible in the near future, especially if the innovation principle is brought fully in line with the evolving data-driven nature of digital innovation and provides more guidance to the Commission on how to design experimental regulation, including inter alia so-called ‘regulatory sandboxes’. Finally, the Commission should ensure that the innovation principle is given prominence with the transition to the Horizon Europe programme, in particular due to the anticipated launch of ‘missions’ in key domains.

Artificial Intelligence (AI) is an area of research and technology application that can have a significant impact on public policies and services in many ways. In just a few years, it is expected that the potential will exist to free up nearly one-third of public servants’ time, allowing them to shift from mundane tasks to high-value work. Governments can also use AI to design better policies and make better decisions, improve communication and engagement with citizens and residents, and improve the speed and quality of public services. While the potential benefits of AI are significant, attaining them is not an easy task. Government use of AI trails that of the private sector; the field is complex and has a steep learning curve; and the purpose of, and context within, government are unique and present a number of challenges.

This article investigates the role of cost-benefit analysis during an antiregulatory period. The period since 2016 has seen several new developments, including the first vigorous use by Congress of its power to overturn recently issued regulations and the creation of novel deregulatory mechanisms layered on top of cost-benefit analysis. This period also contains important examples of sharply reversed CBAs, in which regulations that were said to have large net benefits under Obama are instead said to have net costs under Trump. The Trump Administration’s regulatory review initiatives focus heavily on costs, with limited attention to benefits. Case studies of three high-profile regulations show that the economic analysis of one is seriously defective, another admits to having severe limitations, and a third systematically reduces the scale of benefits. Some of these characteristics may be analytically defensible, others seemingly are not. It is even harder to connect Congress’s recent uses of the Congressional Review Act to either a concern about net benefits or even a desire to reduce the economic burdens of regulation. Thus, cost-benefit analysis seems overall a marginal part of current regulatory policymaking.