Introduction: Generating Governance - An Essay Series on Strategies and Challenges in AI Regulation

In this essay series, the authors explore different aspects of emerging AI governance regimes.  Though about quite different topics, the essays have many common threads. Several of the essays demonstrate that many of the difficulties with AI governance are less challenges of AI than challenges of governance generally—navigating power struggles and competing interests, getting buy-in, and avoiding capture. A second theme is the importance of regulatory diffusion and how governance regimes implicitly rely on it. Several of the authors see a vision of diffusion as central to the form of regulation chosen by each government—particularly in the EU’s case, where the government relies on the so-called “Brussels Effect” to assert greater control over global technology and market regulation. Finally, several of the authors examine the EU’s draft AI Act, probably the most globally important pending AI regulation. Specifically, they discuss the impacts of its core focus on risk, an approach that relies heavily on precaution, internal assessments, and enterprise risk management and mitigation, rather than accountability and liability from an oversight body. The choice to use risk management—a choice that the EU is certainly not alone in—has specific consequences that are worth teasing out, and several of the essays here do just that.

About the Author

Andrew Selbst is an Assistant Professor at the UCLA School of Law.