How the AI arms race has become a war for talent

Google DeepMind is losing ground in the intensifying battle for elite AI talent as OpenAI and Anthropic rapidly expand and attract researchers with enormous compensation packages and valuable equity, Fortune reports.

New analysis from Zeki Data shows DeepMind’s share of research and advanced-engineering hires across Europe, the Middle East and Africa fell sharply from 49% in 2022-23 to 18.6% in 2025-26. Globally, DeepMind is still hiring more researchers and advanced engineers than it loses, but its arrivals-to-departures ratio dropped from about 12-to-1 in the second quarter of 2023 to roughly 2-to-1 in the third quarter of 2026. Anthropic’s ratio was 22-to-1.

The shift comes amid a series of high-profile departures involving researchers associated with AlphaGo and AlphaFold. Over the past year, 25% of departing DeepMind researchers went to Anthropic, 21% to Meta and 14% to OpenAI.

Former employees say DeepMind’s changing identity is partly responsible. As Google has focused more heavily on developing and commercializing Gemini, some researchers feel the lab has moved away from the open-ended, long-term science that historically attracted academics. Tighter publication rules for commercially sensitive research may have further weakened that appeal.

The talent gap is reflected in annualized workforce growth. Since 2022, research and engineering headcount has grown about 97% annually at OpenAI, 152% at Anthropic and 27% at DeepMind.

DeepMind is gaining talent in robotics and AI for science, but losing ground in large language models, multimodal AI and computer vision, demonstrating how talent retention has become a critical competitive battleground in the AI industry.

Fortune has the full story. A subscription may be required.