Alexander C. Kurtz
Abstract: Generative artificial intelligence (AI) is a nascent technology that threatens great harm while simultaneously promising significant benefits. Although AI possesses incredible capabilities to solve problems and complete tasks, it also poses two major threats: technological dislocation and existential catastrophe. Accordingly, the novelty and power of generative AI has led to calls for regulation, including requests for a new federal agency. This Article examines whether Congress should authorize a new federal agency for AI and, if so, what its scope of authority should be. In contemplating such regulation, it is important to consider whether existing agencies can adequately address AI’s primary threats, how a new agency could be helpful, and what costs the creation of a new agency would impose. Weighing these factors, this Article argues that a new, independent federal research agency (without regulatory authority) would adequately address the unmet threats from AI without incurring disproportionate costs.