DeepMind CEO proposes FINRA-style standards body to regulate frontier AI model releases

DeepMind CEO Demis Hassabis called for the creation of an independent AI standards body modeled after the Financial Industry Regulatory Authority (FINRA) to oversee the release of frontier AI models. In an X post titled 'A Framework for Frontier AI and the Dawning of a New Age,' Hassabis argued that such a body could test frontier models and develop best practices for their deployment. The proposal suggests that initially, frontier labs would voluntarily share models with the standards body for review up to 30 days before release.

Hassabis outlined a phased approach where, if the assessment protocol proves effective and robust, formalization could follow, making it mandatory for frontier models to pass the regulator’s review to be deployed in the U.S. market. The standards body would also collaborate with labs to address critical post-release vulnerabilities, ensuring ongoing oversight beyond initial deployment.

The proposed regulator would be backed by the U.S. government but funded by the AI industry, operating independently to mitigate conflicts of interest. Hassabis envisions staffing the body with technical experts from within the industry, open-source representatives, and potentially outsourcing evaluations to specialized AI safety groups. This structure aims to ensure the regulator remains technically focused while supporting innovation and responsible behavior.

The proposal comes amid criticism of existing ad hoc government reviews, such as those conducted for Anthropic’s Mythos and OpenAI’s Sol, which were criticized for lacking technical expertise and opaque decision-making processes regarding model releases. Hassabis’s framework seeks to address these concerns by institutionalizing a transparent, technically rigorous review process.

The idea of AI regulation remains contentious, with White House AI advisor and a16z general partner Sriram Krishnan recently stating that 'there will not be an FDA for AI.' Hassabis’s proposal positions the standards body as a self-regulatory organization, potentially offering a politically viable path to oversight that avoids direct government bureaucracy.