Designing a FINRA for Frontier AI | Lawfare

The authors contend that this entity should replace the current ad hoc approach to AI regulation. Rather than relying on case-by-case actions by different parts of the government, the article proposes a standing institution with responsibility for overseeing advanced AI development. The proposal is framed as a more systematic alternative to the improvisation that has characterized recent U.S. policy.

The article cites two examples of what it describes as improvised governance. The first is export controls imposed by the Commerce Department on Anthropic's Fable 5. The second is pressure exerted by the White House on OpenAI. The article says these measures show that neither developers nor the government believes the current approach is working. It describes such interventions as responses to particular circumstances rather than parts of a deliberate framework. These examples are presented as evidence that the existing ad hoc system is not producing satisfactory results for either side.

The proposal draws on the FINRA model. The article suggests that a similar body, placed under federal supervision, could provide clearer oversight for frontier AI development. It argues that the pace of AI advancement requires a more permanent and predictable mechanism than export controls or White House pressure. The piece is hosted by Lawfare, which focuses on national security law and policy.