Why do AI experts prefer Anthropic and OpenAI over Google?
The AI job market has become highly competitive, with top talent increasingly favoring OpenAI and Anthropic over Google DeepMind. New data from Zeki Data indicates that Google's research lab is struggling to retain its leading researchers and engineers.
The job market in the field of artificial intelligence has resembled a professional sports league in recent years, with huge salaries and stock packages that turn employees into millionaires. However, the research lab Google DeepMind, which previously led the market, now finds itself on the losing side. New data from the research firm Zeki Data shows that while competitors OpenAI and Anthropic are making significant achievements in recruiting top researchers and engineers, Google and Meta are struggling to maintain their status.
The numbers behind the employee exodus
According to the data, DeepMind's market share in recruiting researchers and engineers in Europe, the Middle East, and Africa dropped from 49% in 2022-2023 to just 18.6% in 2025-2026. The ratio of hires to departures at the company dropped from 12 to 1 in 2023 to just two hires for every employee leaving in 2026. For comparison, at Anthropic, the ratio stands at 22 hires for every departure. Of the employees who left DeepMind in the last year, a quarter moved to Anthropic and 14% to OpenAI.
According to the report, the change stems mainly from a shift in DeepMind's identity. While in the past the company attracted academics looking for free research, it is now focused on the commercial development of the Gemini model to close the gap with competitors. In addition, the company tightened rules on publishing research and imposed a six-month delay on the publication of sensitive developments to prevent competitors from benefiting from them. Long-time employees report frustration that they are required to build commercial products instead of engaging in pure research.
A wave of departures of top scientists
The result is a wave of departures of key figures. This month, long-time chief scientist Jeff Dean and other senior researchers left to start a new startup. At the same time, CEO Demis Hassabis stepped down from day-to-day management and moved to the role of chairman. Another notable departure is that of reinforcement learning pioneer David Silver, who founded a startup now valued at 5.1 billion dollars after raising a 1.1 billion dollar seed round.
Joining Silver were also Noam Shazeer, one of the leaders of the Gemini project who moved to OpenAI, and John Jumper, who shared this year's Nobel Prize in Chemistry for the development of AlphaFold and moved to Anthropic. Although DeepMind still has the computing power, reputation, and money of Google, it seems that in a market where the best researchers can choose where to work and what to work on, its historical advantages are eroding.
