Mark RadarMARK RADAR
About
EN
Sign in

Google’s Top AI Talent Continues to Depart for Anthropic and OpenAI

1 reports · First detected 2026-06-25 · Last active 2026-06-25

Google is competing with Anthropic and OpenAI for leadership in generative AI, with the development of Gemini and Claude relying on core researchers. In 2024, Google brought Character.AI co-founder Noam Shazeer back through a licensing deal worth about $2.7 billion. The latest departures underscore that heavy investment may not secure talent over the long term.

On June 24, 2026, Gemini researchers Jonas Adler and Alexander Pritzel were set to leave Google for Anthropic. Shazeer announced on June 18 that he would join OpenAI, while 2024 Nobel laureate in chemistry John Jumper said on June 19 that he would join Anthropic. In all, four key figures left within a week.

All Coverage

1 original reports

The Backstory

The history behind this event
DeepMind Loses Top AI Talent to Anthropic and OpenAI2026-08-27 · 1 reports · similarity 0.86

Google DeepMind has long been one of artificial intelligence’s premier research institutions, backed by Google’s computing infrastructure and financial resources. But its growing emphasis on commercializing Gemini has raised concerns that the open-ended research culture that attracted leading scientists is being diluted. Retaining elite researchers is increasingly critical as DeepMind competes with well-funded rivals for the people capable of advancing frontier AI systems.

The latest recruitment data show a sharp decline in DeepMind’s share of AI hiring and in its ratio of incoming to departing talent. Several prominent researchers and senior executives have recently moved to Anthropic or OpenAI, while others have left to start companies. The pattern suggests that departures are becoming a broader organizational challenge rather than isolated moves, steadily eroding DeepMind’s advantage despite its continued access to substantial capital and computing power.

Google Loses Top AI Researchers as Compute Strains and Bureaucracy Bite2026-08-09 · 2 reports · similarity 0.88

Google is struggling to retain elite artificial-intelligence researchers as competition for scientists capable of building frontier models intensifies. Authors of influential papers and other key personnel have moved to OpenAI and Anthropic, raising questions about the depth and continuity of the teams behind Gemini. Their departures matter because cutting-edge AI development depends not only on capital and chips, but also on a relatively small pool of researchers with experience training and scaling advanced models.

Recent reports identify uneven access to Google’s in-house TPU computing capacity and lengthy, rigid approval procedures as major sources of frustration. Internal resource competition is said to favor businesses that commercialize computing power, constraining some Gemini research and slowing experiments. That has helped drive key researchers toward OpenAI and Anthropic, where faster access to compute and more flexible decision-making may offer an advantage. The reports did not disclose specific financial amounts or departure dates.

Mark Radar|MARK RADAR

If you search news on Google, you can set Mark Radar as a preferred source—our coverage will show up more often in your results. Set as preferred source on Google →

All times are in Taipei time (GMT+8)