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

The proposed body would have the authority to screen the world's most advanced AI models and, if necessary, coordinate an industry-wide slowdown in development to mitigate potential dangers.

Hassabis suggests that the watchdog could impose temporary halts or safety requirements on labs that fail to meet certain thresholds.

The proposal comes amid ongoing global debates over AI regulation. has yet to pass comprehensive federal legislation.