Demis Hassabis proposed IAEA-style AI watchdog as Google DeepMind CEO
Just before stepping down as CEO of Google DeepMind, Demis Hassabis was working on an idea for a new, independent group to keep AI safe.
Inspired by the International Atomic Energy Agency, his plan was for this group to set safety standards and watch over the development of artificial general intelligence, or AGI.
Hassabis shifts focus to scientific breakthroughs
Hassabis talked with government officials and tech leaders about making this happen.
In July 2026, he publicly suggested that the industry should fund a group (kind of like how finance has its own watchdog) to check powerful AI models.
Now, shifting roles to chairman at DeepMind and chief scientist at Alphabet, he's focusing more on scientific breakthroughs.
His push for better AI oversight matches what other big names in tech are calling for, like OpenAI's Sam Altman wanting global rules and Anthropic suggesting we slow things down worldwide.