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

What happened

Demis Hassabis, CEO of Google DeepMind and Nobel laureate, published a manifesto on July 14, 2026 titled 'A Framework for Frontier AI and the Dawning of a New Age,' calling on the US to lead a new international AI regulatory body.

Context and impact

Hassabis says current AI-driven cyber risks are 'warning shots' — within 18 months, dangerous biological and nuclear capabilities could be embedded in open-source models beyond any government's reach. The manifesto comes as Hassabis believes AGI is 'probably just a few short years away.'

Details

• Modeled on FINRA: privately funded, industry-staffed, answerable to US government
• Frontier labs: voluntary model submission up to 30 days before public release
• Testing scope: dangerous cyber, biological, and deception capabilities
• Board: majority independent (Turing Award winners, domain experts, industry and government)
• Goal: operational before end of 2026
• Hassabis: current AI-driven cyber threats are 'warning shots'