Google DeepMind Loses Ground in Elite AI Talent as Rivals Close the Gap

Google DeepMind is seeing a significant decline in its share of elite artificial intelligence talent as rival AI laboratories intensify recruitment across key global markets, according to a report by Fortune.

An analysis by UK-based intelligence firm Zeki Data found that DeepMind’s share of research and advanced-engineering hires across Europe, the Middle East and Africa (EMEA) fell from 49% in 2022–23 to 18.6% in 2025–26. The report described the decline as the steepest recorded for a major AI laboratory in any region.

Rivals Gain Ground in Talent Race

The shift comes as OpenAI, Anthropic and Meta expand their AI teams and compete aggressively for experienced researchers and engineers.

DeepMind continues to hire more people than it loses, but the gap between arrivals and departures has narrowed considerably. According to Zeki Data, its arrival-to-departure ratio fell from approximately 12-to-1 in Q2 2023 to around 2-to-1 in Q3 2026.

The movement of departing talent has also benefited competitors. Among researchers and advanced engineers who left DeepMind over the past year, 25% joined Anthropic, 21% moved to Meta and 14% transitioned to OpenAI.

Competitor laboratories recorded stronger recruitment ratios during the same period, with Anthropic at 22-to-1 and OpenAI at 5.7-to-1, according to the analysis.

High-Profile Departures

The decline in DeepMind’s talent share follows several prominent departures from the company’s scientific and research ranks.

Long-serving Chief Scientist Jeff Dean left after 27 years to co-found startup Discovery Loop alongside former colleagues Sanjay Ghemawat, Oriol Vinyals and Quoc Le.

Other notable departures include Noam Shazeer, a co-lead on Gemini who moved to OpenAI, and John Jumper, the 2024 Chemistry Nobel laureate who co-led AlphaFold and subsequently joined Anthropic.

Jumper’s move to Anthropic was accompanied by key researchers Jonas Adler and Alexander Pritzel. Overall, Zeki’s analysis found that 13 of the 29 named authors of the seminal AlphaFold2 research paper have now left DeepMind.

Pressure on Foundational AI Research

The talent movement has been particularly pronounced in some of DeepMind’s foundational AI research areas.

Zeki found that 19.2% of recent departures specialised in large language models and multimodal systems, compared with 15.6% of new hires in those categories. The company has also experienced notable losses in computer vision.

DeepMind has, however, recorded net talent gains in selected specialist areas. Its workforce has expanded in robotics, embodied AI and machine-learning applications for scientific discovery, suggesting that recruitment remains active in areas aligned with the laboratory’s broader research priorities.

Changing Research Environment

Internal policy changes and intensifying competition for AI specialists have also been cited as factors behind employee attrition.

According to a Financial Times report referenced in the material, DeepMind introduced stricter internal peer-review processes and a six-month publication embargo for strategically sensitive generative-AI research papers. The company defended the measures as necessary to protect intellectual property while continuing to contribute to academic research.

Current and former employees cited in the reporting pointed to changes in research culture, aggressive equity packages offered by pre-IPO AI companies and concerns around morale amid leadership restructuring.

The leadership changes include Demis Hassabis moving into a chairman role, while CTO Koray Kavukcuoglu assumed greater operational control.

As competition for elite AI researchers continues to intensify, the shifting movement of talent highlights the increasingly important role that compensation, research freedom, corporate culture and long-term incentives play in determining where leading AI scientists and engineers choose to work.