Google DeepMind Loses Talent War as Rivals Poach Top AI Researchers
New data shows the lab's European market share plummeting from 49% to 18.6% as OpenAI, Anthropic, and Meta aggressively recruit its scientists.
DeepMind's Talent Dominance Erodes
Google DeepMind, once the premier destination for elite AI researchers in Europe, is losing ground in the intensifying competition for top talent. New data from UK-based Zeki Data, shared exclusively with Fortune, reveals a dramatic reversal: DeepMind's share of research and advanced-engineering hires across Europe, the Middle East, and Africa plummeted from 49% in 2022–23 to just 18.6% in 2025–26—the steepest decline recorded for any major AI lab in the region.
The shift reflects broader turbulence at the storied research organization. This month alone, DeepMind lost Jeff Dean, its chief scientist and 27-year Google veteran, along with senior fellow Sanjay Ghemawat and researchers Oriol Vinyals and Quoc Le, who departed to launch startup Discovery Loop. On the same day, cofounder and CEO Demis Hassabis announced he would step back from daily operations to become chairman and Alphabet's chief scientist, handing control to CTO Koray Kavukcuoglu.
Why It Matters
The talent exodus threatens Google's ability to compete in the AI race at a critical moment. A small cohort of elite researchers drives progress on frontier models—the systems underpinning the current AI boom. Losing these scientists to rivals like OpenAI and Anthropic doesn't just slow DeepMind's own development; it accelerates competitors' capabilities. For a company trying to close the gap with OpenAI's ChatGPT dominance, the hemorrhaging of research talent represents both a strategic and reputational setback.
Where the Researchers Are Going
Anthropichas emerged as the leading destination for departing DeepMind staff, capturing 25% of exits over the past year, according to Zeki's analysis. Meta attracted 21%, while OpenAI drew 14%. The data shows DeepMind's arrivals-to-departures ratio collapsed from roughly 12-to-1 in Q2 2023 to approximately 2-to-1 in Q3 2026. By comparison, Anthropic's ratio stood at 22-to-1 in 2025.
High-profile departures include David Silver, the reinforcement-learning pioneer behind AlphaGo, who left to found London startup Ineffable Intelligence, now valued at $5.1 billion. In June, Google lost Gemini co-lead Noam Shazeer to OpenAI and Nobel laureate John Jumper to Anthropic within 24 hours. Of the 29 authors on the landmark AlphaFold2 paper, 13 have since left DeepMind.
Product Focus Displaces Research Culture
Current and former employees told Fortune the talent drain stems from multiple factors: aggressive poaching by rivals offering cash-heavy compensation, mounting frustration over Google's position in the AI race, declining morale, and the appeal of pre-IPO equity at OpenAI and Anthropic.
But a deeper issue involves DeepMind's shifting identity. As Google has pushed to commercialize its Gemini model and close the gap with rivals, the lab has moved away from the open-ended, long-horizon research that once attracted academics. DeepMind tightened publication rules in 2025, introducing a tougher review process and six-month embargo on strategically sensitive papers—a shift that created friction with researchers who joined to pursue unconstrained, blue-sky work.
"Most old timers who joined DeepMind before the ChatGPT moment, joined to be part of an AI research lab," one former engineer said. "Anyone who joined before 2023 thought they were joining an AI research lab, and suddenly they were asked to build products for Google."
Zeki's analysis, which tracks 20,900 people in research and advanced-engineering roles at 10 companies, found DeepMind suffered net losses in large language models and multimodal systems expertise. The company is gaining ground in robotics and embodied AI, but those gains haven't offset departures in core areas.
Tom Hurd, founder of Zeki Data, noted the pattern: "The more research-heavy people feel that they'd quite like to look at opportunities elsewhere, and then you start to see a much larger number than usual going to set up their own thing and bringing some of their friends with them."
The details were first reported by Fortune.
This is an original analysis by the Omega editorial team. Source reporting: AI Watch.
Want systems like this working for your business?
Book a Call
