The real issue is not which level of government should regulate AI, but whether government institutions possess the knowledge, incentives, and institutional capacity necessary to regulate a rapidly evolving technology in the first place.