Google's AI shakeup: DeepMind's Hassabis steps aside, senior scientists depart

Google spent the early ChatGPT era looking like the industry's slowest kid on the block, scrambling after Microsoft's Copilot launch caught it off guard. Fast forward to now, and the company sits near the top of the AI pile, with Gemini models and DeepMind's research pumping out headlines. But the leadership chart at the top is starting to wobble, and the newest crack is a big one: Demis Hassabis, DeepMind's CEO and one of its original founders, is stepping back from daily operations to take an oversight role at Alphabet.

Hassabis isn't leaving outright, but the shift matters. He's been the public face of DeepMind since Google absorbed the lab, and arguably the loudest voice inside Google insisting that artificial general intelligence — AI that matches or exceeds human-level reasoning across the board — is not some far-off sci-fi milestone but a near-term inevitability. In his memo announcing the move, he doubled down on that stance, writing that he's spent his whole career chasing AGI and now feels it's within reach.

The timing is what makes this notable. Hassabis's step back arrives amid a string of senior departures from Google's AI ranks, a pattern that's hard to wave away as coincidence. When a company is simultaneously claiming to be on the verge of one of technology's biggest breakthroughs and watching senior scientists walk out the door, those two facts sit uneasily next to each other.

Alphabet hasn't spelled out exactly who takes over Hassabis's day-to-day responsibilities, nor detailed what his new oversight role will actually involve beyond vague gestures toward strategy. For a company that's poured enormous resources into convincing the public — and its own shareholders — that it's leading the AGI race, leaving those specifics fuzzy is a curious choice.