In an X post on Tuesday morning, Google DeepMind CEO Demis Hassabis called for the creation of a new regulatory body to oversee frontier model releases. Titled “A framework for frontier AI and the dawn of a new era,” the publication advocates for a “standards body” modeled on the Financial Industry Regulatory Authority (FINRA), which could test frontier models and develop best practices for publication.

“Initially, Frontier Labs would voluntarily share models with the standards body for review up to 30 days before release,” the post reads. “Once the evaluation protocol is proven effective and robust, formalization could follow quickly, meaning it would be required to be approved by Frontier Models to be implemented in the US market. The labs would also work with the standards body to address any critical post-launch vulnerabilities.”

The proposed system would build on ad hoc reviews by the US government of Anthropic’s Mythos and OpenAI’s Sol. Those reviews drew significant criticism for a lack of technical expertise and opaque decision-making about when a model could be released. Under Hassabis’ proposed regulator, those decisions could be handed over to a new organization, backed by the U.S. government but funded by the AI ​​industry and operated independently.

The prospect of AI regulation remains controversial for both the tech industry and the Trump Administration. More recently, Sriram Krishnan, White House AI advisor and general partner at a16z, ruled out the possibility of an AI regulator within the executive branch, saying “there will not be an FDA for AI.”

Establishing the standards body as a self-regulatory organization like FINRA could be one way to address those concerns. Hassabis anticipates that the regulator will have open source representatives and technical experts from the industry, along with the financial backing from AI labs that would be necessary to retain them. They could even outsource some assessments to the growing group of AI security groups that could specialize in specific risks.

“The advantage of this approach is that it would be technically focused and, at the same time, support innovation and incentivize responsible behavior,” says Hassabis. “It is designed to keep pace with the acceleration of the field and adapt to increased risks as they are identified, and could be escalated if the severity of the situation demands it.”