Google DeepMind's Demis Hassabis Calls for US-Led AI Safety Body with Mandatory Pre-Deployment Testing

Google DeepMind CEO Demis Hassabis is calling for a US-led, FINRA-style public-private AI safety body with mandatory pre-deployment testing for cybersecurity, biological, and national security risks, funded primarily by industry, with the power to slow frontier AI development if threats become too severe.

• Google DeepMind's Demis Hassabis is publishing a detailed AI safety framework on Twitter, calling for a US-led, federally overseen public-private standards body modeled after FINRA to evaluate and regulate frontier AI models.
• Hassabis is recommending mandatory pre-deployment testing for cybersecurity, biological threats, and national security risks, with frontier labs required to publish technical documentation, invest in safety, and maintain strong cybersecurity practices.
• Hassabis is proposing that the standards body be funded primarily by industry rather than government to ensure stable resources, while also pushing for international cooperation and the ability to slow down frontier AI development if risks become too severe.