John Jumper, the Google DeepMind scientist best known for co-creating AlphaFold, is leaving the Alphabet-owned AI laboratory to join Anthropic. He announced the decision on June 19, saying he would make the move after nearly nine years at DeepMind, where he serves as a vice president and engineering fellow. The departure moves one of the sector’s most prominent AI-for-science researchers to an independent frontier-model developer competing directly with the world’s largest technology groups.
A Career Defined by AlphaFold
Jumper’s name is closely tied to AlphaFold, the artificial intelligence system developed at DeepMind to predict the three-dimensional structures of proteins from their amino acid sequences. The project addressed a longstanding scientific challenge, and AlphaFold’s predictions have helped researchers examine biological questions that could previously require substantial time and laboratory resources. Alongside DeepMind chief executive Demis Hassabis and American scientist David Baker, Jumper received the 2024 Nobel Prize in Chemistry for work that demonstrated AI’s potential to accelerate scientific research.
DeepMind’s Response
In his announcement, Jumper thanked Hassabis for placing him in charge of the AlphaFold team shortly after he completed his doctorate, and described Google DeepMind as a special place. Hassabis responded publicly by crediting their collaboration with showing what AI could achieve in science and medicine, while a DeepMind spokesperson said the company was grateful for Jumper’s contributions. The statements underline the significance of the relationship while also presenting the departure as an amicable transition rather than a public rupture.
Anthropic’s Opportunity
Anthropic has confirmed that Jumper will join the company, but it has not disclosed his title, responsibilities or start date. That leaves open whether his work will centre on scientific applications, fundamental model research, safety, or a combination of these areas, though his record gives Anthropic a high-profile researcher with rare experience in deploying AI for scientific discovery. The appointment could also broaden external attention on Anthropic’s research ambitions beyond its Claude models and enterprise software products.
A Sharper Talent Contest
Jumper’s exit follows the announced move of Google Gemini co-lead Noam Shazeer to OpenAI only days earlier, drawing renewed attention to the competition for elite AI researchers. Google, Meta, OpenAI and Anthropic are investing heavily in computing infrastructure, model development and recruitment as they seek to build more capable systems and convert research advances into commercial products. For smaller but well-funded AI companies, senior hires with proven research records can strengthen technical capacity, organisational credibility and their ability to attract further talent.
Implications for Google DeepMind
The loss is particularly notable for Google DeepMind because AlphaFold remains one of the clearest examples of the lab translating advanced machine learning into a widely used scientific resource. DeepMind retains substantial research depth and the broader support of Alphabet, but departures of senior leaders can reshape the distribution of expertise and influence across a fast-moving field. The company will now need to demonstrate that its scientific research agenda and wider product strategy can continue attracting and retaining researchers of Jumper’s standing.
Jumper’s move closes an important period at Google DeepMind, where his work helped turn protein-structure prediction into a defining application of modern AI. It also gives Anthropic a scientist whose career connects foundational research with a practical platform used by researchers around the world. As frontier AI companies compete for talent and technical leadership, the transfer illustrates how scientific capability has become central to the industry’s strategic contest.