![](\"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/08\/GettyImages-2276578034-e1787759543733.jpg?w=2048\")
<\/p>\n<\/p>\n
Over the last few years, the AI talent market has looked more like a professional sports draft than a conventional hiring cycle. Cash-rich labs are offering salaries more commonly associated with pro athletes, plus equity packages that can turn early employees into billionaires. The allure of life-changing paychecks has resulted in elite researchers swapping between rival labs at dizzying speed.\u00a0<\/p>\n
Google DeepMind, which was once the leading destination for many of those researchers\u2014especially in Europe\u2014is now finding itself on the losing side of that contest, according to a new overview of the flow of engineering talent exclusively shared with Fortune<\/em>.\u00a0 A new analysis from Zeki Data, UK-based data intelligence company, shows that where OpenAI and Anthropic are making gains in the AI talent market, Google DeepMind and, to some extent, Meta are stumbling.\u00a0<\/p>\n Interviews with several current and former staff suggest DeepMind\u2019s changing identity has weakened part of its historic appeal. As Google has pushed to close the gap with OpenAI and Anthropic, the lab has become more tightly organized around improving and commercializing Gemini\u2014a shift that some researchers see as displacing the open-ended, long-horizon science that once made DeepMind such a desirable destination for academics.<\/p>\n Three current and two former DeepMind staffers told Fortune<\/em> the recent string of departures come down to a mix of factors: rival labs like Meta and Microsoft aggressively poaching talent with cash-heavy offers, mounting frustration inside Google over where it stands in the AI race, sinking morale, and the pull of pre-IPO stock at competitors such as OpenAI and Anthropic.<\/p>\n This month alone, Google DeepMind lost Jeff Dean, the company\u2019s long-time chief scientist and a 27-year veteran, alongside senior fellow Sanjay Ghemawat and researchers Oriol Vinyals and Quoc Le, who left to launch a startup called Discovery Loop. On the same afternoon, Demis Hassabis, DeepMind\u2019s cofounder and chief executive, said he would step back from day-to-day control of the lab to become its chairman and Alphabet\u2019s chief scientist, handing operational control to CTO Koray Kavukcuoglu.<\/p>\n The new data from Zeki suggests the public exits are part of a broader reversal. For example, DeepMind\u2019s share of research and advanced-engineering hires across Europe, the Middle East and Africa fell from 49% in 2022\u201323 to 18.6% in 2025\u201326\u2014the sharpest market-share drop that Zeki recorded for a major AI lab in any region.<\/p>\n \u201cThey had the crown in Europe forever, and then it started to erode from a very high base,\u201d Tom Hurd, founder of Zeki Data, told Fortune<\/em>. \u201cThe likes of Microsoft AI Superintelligence and Meta Superintelligence are eating into their market share, and then there\u2019s OpenAI and Anthropic on the side.\u201d<\/p>\n The stakes of recruiting and retaining elite talent are high in the current hypercompetitive AI market. A relatively small group of researchers and engineers have the experience to train and improve the frontier models driving the AI boom. Their work can determine how quickly a lab improves its models, whether it can turn research breakthroughs into products, and how credibly it can attract the next wave of talent. Hiring and retaining these top engineers has proved difficult over the last few years, even for the industry\u2019s best-funded companies.<\/p>\n Globally, DeepMind is still bringing in more research and advanced-engineering staff than it is losing. But its arrivals-to-departures ratio\u2014a measure of hires relative to exits\u2014has fallen sharply, from about 12-to-1 in the second quarter of 2023 to roughly 2-to-1 in the third quarter of 2026, according to Zeki\u2019s data. That means it is now adding about two people in these roles for every one who leaves, rather than roughly 12. Comparatively, Meta\u2019s ratio in 2025 was 3 to 1, OpenAI\u2019s was 5.7 to 1, and Anthropic\u2019s was 22 to 1, per the report.\u00a0<\/p>\n Anthropic has become the leading destination for departing DeepMind researchers and advanced engineers. Of the people who left DeepMind in the past 12 months, 25% went to Anthropic, 21% to Meta, and 14% to OpenAI, according to Zeki.<\/p>\n Representatives for Google DeepMind did not respond to a request for comment on Zeki\u2019s findings.\u00a0<\/p>\n The deterioration in DeepMind\u2019s talent flows coincided with the lab tightening its publication rules, Hurd said, something that Zeki researchers say may have weakened one of the lab\u2019s most important draws for research-minded staff. The Financial Times <\/em>first reported in April of 2025 that DeepMind had introduced a tougher internal review process and a six-month embargo for some strategically sensitive generative-AI papers, as the company sought to prevent competitors from benefiting from its research.<\/p>\n The FT<\/em> reported that the new approach made it harder for researchers to publish studies, particularly work that could expose product weaknesses or reveal commercially valuable techniques. DeepMind said at the time that it remained committed to research publication and was updating its policies to preserve its teams\u2019 ability to contribute to the broader research ecosystem.<\/p>\n For a lab that built its reputation on public breakthroughs such as AlphaGo and AlphaFold, the shift created friction with researchers who had joined to pursue relatively unconstrained, blue-sky work.\u00a0<\/p>\n \u201cMost old timers who joined DeepMind before the ChatGPT moment, joined to be part of an AI research lab,\u201d one former DeepMind engineer told Fortune<\/em>. \u201cAnyone who joined before 2023 thought they were joining an AI research lab, and suddenly they were asked to build products for Google.\u201d<\/p>\n The firm found that research and engineering hiring across the sector has grown at a compound annual rate of 23% since 2022. But the growth is flowing disproportionately to newer frontier labs. OpenAI\u2019s research and engineering headcount has grown at roughly 97% annually and Anthropic\u2019s at 152%, compared with 27% for Google DeepMind. OpenAI and Anthropic are starting from a much smaller base, but the figures show how quickly their organizations have expanded relative to DeepMind.<\/p>\n DeepMind is also encountering new competition in regions it long dominated. Mistral AI and Anthropic have been the primary beneficiaries of DeepMind\u2019s declining share in Europe, the Middle East and Africa, according to Zeki. In Asia-Pacific, where DeepMind opened a new research lab in Singapore last November, domestic players including ByteDance, Sakana AI and Sarvam AI are building share in markets the incumbent labs historically underinvested in.<\/p>\n This shift is also visible in a growing list of prominent departures.<\/p>\n David Silver, the reinforcement-learning pioneer behind AlphaGo, AlphaZero and AlphaStar, left after nearly 13 years at DeepMind to launch Ineffable Intelligence earlier this year, a London startup now valued at $5.1 billion after a $1.1 billion seed round. Other long-serving DeepMind employees who departed include\u00a0 Wojciech Czarnecki, now chief technology officer at Fundamental, and Lasse Espeholt.<\/p>\n Misha Laskin and Ioannis Antonoglou, the latter of whom worked directly with Silver on AlphaGo, also left in 2024 to found Reflection AI.<\/p>\n This summer also saw another sharp run of exits. In June, Google lost Gemini co-lead Noam Shazeer to OpenAI and John Jumper, who shared the 2024 Nobel Prize in Chemistry for AlphaFold with Hassabis, to Anthropic within a day of one another. Fellow AlphaFold researchers Jonas Adler and Alexander Pritzel followed Jumper to Anthropic shortly afterward.<\/p>\n \u201cThe more research-heavy people feel that they\u2019d 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,\u201d Hurd said. \u201cOver time, that\u2019s added up to a large number of people.\u201d<\/p>\n Zeki\u2019s data suggests the losses are not evenly spread. Comparing the expertise of everyone who left DeepMind with those who joined during the past 12 months, the firm found a net loss in large language models and multimodal systems: 19.2% of leavers specialized in those areas, compared with 15.6% of joiners. The company has also lost ground in computer vision, per the report.\u00a0<\/p>\n The AlphaFold team\u2014defined as the group that produced AlphaFold2, the AI system that predicts the three-dimensional structure of proteins from their genetic sequence\u2014has also suffered heavy losses. Of the 29 named authors on the AlphaFold2 paper, 13 have left DeepMind since, according to Zeki. The Financial Times <\/em>reported in July that DeepMind had reassigned most of the original paper\u2019s authors during the preceding year, with staff moving to Gemini-related projects, as well as enzyme design, genomics, nuclear fusion and Isomorphic Labs, Alphabet\u2019s drug-discovery subsidiary.<\/p>\n DeepMind is gaining ground, however, in robotics, embodied AI and machine learning for science. Those fields reflect areas where the company is expanding its capacity\u2014and, in some cases, competing for talent with Nvidia as much as with rival AI labs.<\/p>\n Zeki\u2019s analysis tracks 20,900 people in research and advanced-engineering roles at 10 companies, using publicly available information compiled in August 2026. The group includes research scientists, research engineers, machine-learning and deep-learning engineers, applied scientists and members of technical staff; it excludes managers, executives, interns, customer-facing roles and non-technical staff. Because the analysis relies on public information, it may undercount the total number of people in these roles.<\/p>\n DeepMind still has Google\u2019s compute, cash, reputation, and institutional reach to attract staff. But in a market where the best researchers can choose almost anywhere, and work on almost anything, some of its old advantages appear to be harder to preserve.<\/p>\n<\/div>\n #Google #DeepMind #losing #grip #elite #talent<\/p>\n","protected":false},"excerpt":{"rendered":" Over the last few years, the A… <\/p>\n","protected":false},"author":1,"featured_media":3266,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[428,5946,2195,5948,636,5947,4237,480,571],"class_list":["post-3265","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-finance-news","tag-anthropic","tag-battle-for-talent","tag-deepmind","tag-elite","tag-google","tag-grip","tag-losing","tag-openai","tag-talent"],"yoast_head":"\nA string of high-profile exits<\/strong><\/h2>\n
Language model and science teams lose ground<\/strong><\/h2>\n