AI, Expertise, and Administrative Law • News & Events • Penn Carey Law

Professor Cary Coglianese explores the limitations of artificial intelligence (AI) in federal agency decision-making in an essay for the Yale Journal on Regulation’s online publication, Notice & Comment.
In the essay titled “AI, Taxi Drivers, and Administrative Law,” Coglianese cautions against federal agency officials solely relying on AI for policy recommendations, emphasizing the importance of reasoned judgment.
He uses the analogy of an overconfident taxi driver to highlight the risks of AI overconfidence and asserts that the Administrative Procedure Act (APA) serves as a critical safeguard against relying too heavily on AI.
Coglianese acknowledges the utility of AI in supporting decision-making but stresses that it should not replace pivotal policy judgments.
He emphasizes that regulatory decisions under the APA need to be grounded in more substantial reasoning than generalized AI opinions.
An edited version of the essay appeared in The Regulatory Review titled “Administrative Law and AI’s Overconfidence.”
The full essay can be read in its original form in the Yale Journal on Regulation.

You may have missed