Exclusive: Google's Hassabis calls for U.S.-led global AI watchdog

Demis Hassabis, the Nobel Prize-winning CEO of Google DeepMind, is calling for a U.S.-led global AI watchdog to screen the world's most powerful models before they're released, according to Axios. Publishing a manifesto Tuesday, Hassabis proposes a regulatory body modeled on the Financial Industry Regulatory Authority that would test frontier models for dangerous capabilities in cyber security, biology, and deception — starting with voluntary industry participation and potentially moving to mandatory approval for U.S. deployment. Hassabis has been quietly building support for the plan among the Trump administration, rival lab leaders, and European officials, and says other major AI labs agree this is where the industry needs to head. The timeline is aggressive: he's hoping the watchdog could be operational before year-end, as he warns that advanced AI risks — from cyber threats to biological and nuclear dangers — could escape government oversight within eighteen months if left unaddressed.

Listen to this story

Hear this and more stories in a personalized audio briefing.

Open The Chonkerton