Google DeepMind's AlphaFold Lead John Jumper Joins Anthropic

John Jumper, the AI researcher who led the AlphaFold team at Google DeepMind and shared the 2024 Nobel Prize in Chemistry with CEO Demis Hassabis, has announced he is leaving after nearly nine years to join Anthropic. Jumper shared the news on X, and Hassabis responded publicly on X, thanking Jumper for their "extraordinary partnership" and stating that AlphaFold had "changed the world."

The departure carries more symbolic weight than a typical senior hire. AlphaFold's protein-structure prediction became the canonical example of AI accomplishing something scientists had considered decades away, and a Nobel Prize is not a credential most AI labs can claim at all. Losing the person most publicly associated with that achievement, to a direct competitor, is the kind of signal that is hard for an organization to explain away.

This is the third high-profile exit from Google's AI division in a short period, as The Decoder reported. Gemini co-lead Noam Shazeer has joined OpenAI, and AlphaGo and AlphaZero researcher David Silver departed to launch his own startup focused on world models and reinforcement learning. Three named departures in rapid succession is a pattern, not a coincidence, and it suggests structural issues with retention rather than just the pull of individual opportunities.

What the reporting does not give you is what Jumper will actually work on at Anthropic, or whether his exit affects AlphaFold's ongoing development at DeepMind. On the competitive picture: insider reports suggesting Gemini 3.5 Pro may struggle against Anthropic and OpenAI at its late-June launch are unverified speculation rather than benchmarked results, so treat those as industry sentiment rather than confirmed data.

For Anthropic, the hire bolsters its credentials as a destination for researchers doing foundational science, the kind of credibility that is hard to buy and relatively easy to build on. For Google, the accumulating departures create pressure to address whatever is driving them before the exits become self-reinforcing.