Summary: John Jumper, co-recipient of the 2024 Nobel Prize in Chemistry, announced on June 19, 2026 that he was leaving Google DeepMind after nearly nine years for Anthropic. The same week, Noam Shazeer announced a move from Google to OpenAI. Jumper co-led AlphaFold, whose protein-structure work earned the Nobel and became infrastructure for drug discovery and molecular biology.
John Jumper — who shares the 2024 Nobel Prize in Chemistry — took to X on Friday, June 19, 2026, to announce that his nearly nine-year run at Google DeepMind is over: he is off to rival Anthropic. TechCrunch covered the news on June 20. Nor was he the only marquee researcher walking out of Google that week — three days earlier, Noam Shazeer, co-author of the Transformer paper and co-lead of Gemini, announced a departure of his own, bound for OpenAI.
A Near-Decade at DeepMind Draws to an End
The note Jumper posted to mark his exit struck a restrained, warm tone. He recalled how DeepMind CEO Demis Hassabis “took a real chance letting me lead the AlphaFold team just six months after finishing my PhD,” and said that “the entire GDM team taught me so much about how to do great science.” His closing line: “GDM is a special place, and I’ll still be excited to hear about what amazing things they discover next.”
Measured against a job title, DeepMind is losing far more than one senior name. Jumper has been the public face of the lab’s “AI for science” line of work, and the address he picks doubles as a referendum on which labs deserve a top researcher’s next decade.
AlphaFold and a Nobel in Chemistry
The 2024 Nobel Prize in Chemistry went jointly to Jumper and Demis Hassabis for AlphaFold, an AI system that takes a protein’s amino acid sequence and predicts its three-dimensional structure. Work that once ranked among structural biology’s most time-consuming and costly experimental problems became all but instantaneous computation — and “AI for science” went from slogan to a track record crowned with a Nobel. In the years since, AlphaFold’s predicted structures have settled into everyday infrastructure for drug discovery and molecular biology, which is why the system’s co-creator commands a stature few AI researchers anywhere can rival.
Bloomberg’s coverage adds one more revealing thread: over the last few years, Jumper has also held a central spot on the Google team that builds coding tools — a corner of the market where Google has come up short with enterprises, and exactly the line where Anthropic is strongest. Seen in that light, where he ended up looks chosen on purpose: science and coding happen to be the two directions in which Anthropic is pushing outward hardest.
That Same Week, Shazeer Heads for OpenAI
The calendar sharpens the drama. On Wednesday, June 17, Character AI co-founder Noam Shazeer announced he too was leaving Google. His time at the company spans more than two decades — minus a three-year Character AI stint that ended when Google reeled him and his team back in with a $2.7 billion deal in 2024 — and he co-authored the 2017 paper “Attention Is All You Need,” the architecture on which modern generative AI rests.
By TechCrunch’s account, OpenAI — sprinting toward its IPO — is bringing him in, and Shazeer is not the only new arrival: the same reporting also has OpenAI hiring Dean Ball, formerly a White House AI policy official, to head a newly formed Strategic Futures team whose brief spans catastrophic risk, recursive self-improvement, and the links between frontier labs and governments. Inside a single week, Google DeepMind parted ways with both a scientist holding a Nobel and one of the Transformer’s inventors, and the ones gathering them up were its two most direct competitors.
Reading the Talent Flow Against the Frontier Race
This wave rewards a layered read. For Anthropic, bringing Jumper aboard strengthens the “AI for science” storyline and keeps alive the recent pattern of landing heavyweight researchers. For OpenAI, Shazeer signing on hands the research story it tells ahead of its IPO one of its most consequential names yet. For Google, this is a fresh warning that top research talent’s allegiance is being revalued by the frontier race — revalued through pay packages, equity, and the latitude to pursue a chosen research agenda. And for everyone else building AI products, the message is that the people behind the field’s foundational techniques are still on the board — and can still be moved.
From here, watch two things. One is the way Hassabis reshapes who leads AlphaFold and the science organization around it — Jumper had run that group since before the Nobel, and the handoff will shape what DeepMind’s science turns into. The other is where Jumper’s day-to-day at Anthropic actually sits: scientific research, or coding products. Whichever way those land, the third week of June 2026 has secured its footnote in the history of AI talent migration.