In a recent Lawfare episode recorded at Columbia Law, Chloé Bakalar of the Responsible Superintelligence Institute and University College London joined Kevin Frazier of the University of Texas School of Law to ask whether constitutional traditions can inform how we govern AI. The conversation treats constitutions as long-lived frameworks that have dealt with power, rights, and institutional checks — potential templates for structuring AI oversight.
Bakalar and Frazier both seem to see value in borrowing from constitutional design, particularly around creating institutions that can adapt while preserving core commitments. However, they appear to diverge on how literally those lessons should be applied. Frazier, who directs an AI innovation and law program, is more open to using constitutional thinking as a direct blueprint, whereas Bakalar, whose work focuses on superintelligent AI risks, cautions that the speed and scale of AI change may outpace constitutional mechanisms built for slower societies.
The episode does not offer a single verdict, but it frames a core tension: constitutions are meant to be durable, yet AI governance may need to be highly responsive. The speakers explore whether a hybrid approach — constitutional principles for values and institutional checks, plus more agile regulatory tools — could work. No concrete proposal emerges, but the discussion highlights that the debate over AI governance is increasingly looking to older political traditions for guidance.