Translatlantic dialogue on AI and Regulation. Book presentation: The EU Artificial Intelligence Act and the Public Sector (eds. J. Ponce / A. Cerillo-i-Martinez)

During the presentation of this book, two of its authors, Professors Coglianese and Rangone, offered us a conversation on AI from the American and European perspectives.

Professor Coglianese spoke first. Professor Coglianese is the Edward B. Shils Professor of Law and Professor of Political Science and the founding Director of the Penn Program on Regulation at the University of Pennsylvania. He specialises in the study of administrative law and regulatory policy, with an emphasis on the design and evaluation of alternative processes and strategies and the role of public participation, technology, and business government relations in policymaking. He is the author of more than 300 articles, book chapters, reports, and essays on a broad range of administrative law and regulatory policy issues. He served as a founding editor of the international peer-reviewed journal, Regulation & Governance. He also founded and serves as the faculty advisor of The Regulatory Review, a global online publication that covers a full range of regulatory law and policy issues. He has chaired and served on National Academy of Sciences committees as well as committees of the American Bar Association’s Section on Administrative Law and Regulatory Policy. A Senior Fellow of the Administrative Conference of the United States (ACUS), he has served as an appointed Public Member of ACUS and as the chair of ACUS’s Rulemaking Committee. He is a member of the American Law Institute and a fellow of the National Academy of Public Administration.

After noting the increase in the use of AI at the federal level in recent years, Coglianese recalled the reasons for using AI based on human fallibilities (physical limitations, bias, group decision-making problems) as well as the positive possibilities (including accuracy, capacity, speed, consistency) and negative ones (such as bias and hallucinations).

Coglianese raised the need to regulate AI but not in form of regulatory guardrails but as regulatory  leashes, “which are flexible and adaptable—just as physical leashes used when walking a dog through a neighborhood allow for a range of movement and exploration. But just as a physical leash only protects others when a human retains a firm grip on the handle, the kind of leashes that should be deployed for AI will also demand human oversight”. He also made a call to consider the role of empathy as a relevant element in the coming years.

Following Professor Coglianese´s intervention, Professor Rangone continued the conversation. She raised the main criticisms of the European rights-based approach to AI: the substitution of investment for regulation, lack of coherence in AI industrial policy, excessive regulation and negative impact on innovation, brain drain to the US or China, and the inability of regulation to govern a changing reality that is ill-suited to the timeframes for approving regulations.

Professor Rangone highlighted how the risk-based model has the problem of quickly becoming outdated, leaving new uses off the regulatory radar. She also pointed out how the regulation of generative AI is rigid, while there are various different functionalities and risks involved.

The results can be excessive regulation, which generates uncertainty and negatively affects the rule of law.

In this context, the proposal of the Digital Omnibus on AI Regulation Proposal has emerged, which the professor considers to be a remedy worse than the disease: it gives more power to big tech, creates legal uncertainty by delaying the entry into force of the AI Act for 18 months, with unclear rules contained in three articles that modify 30 articles of the AI Act and add two new articles. In addition, significant changes are being introduced without citizen participation or impact assessment, invoking urgency and competition needs.

Along with this, Professor Rangone highlighted the challenge posed by the national implementation of the AI Act. It requires necessary European vertical coordination, as is also the case in the US between the federal and state levels, where more than 1,000 regulations are being processed, as well as horizontal coordination of rules and regulators between EU member states.

In this horizontal coordination, it must be taken into account that European models vary, with some member states using bodies independent of government influence (Lithuania, Cyprus, Slovenia, Latvia, and Luxembourg) and others not, such as Italy, Spain, and Ireland, with non-independent administrations.

According to professor Rangone´s point of view, crucial issues have to do with this coordination and the real effectiveness of the approved rules.

In the subsequent dialogue with both speakers, issues such as the ability of the law to cover all the issues arising from the use of AI and the US experience in relation to environmental regulation and lobbying were raised, which may help to understand what is happening in Europe with regard to AI (from https://rednmr.wordpress.com/blog/)

The full record of the event is available here.