John Jumper, a VP-level engineer at Google DeepMind and co-lead of the AlphaFold protein-folding project, has left the company to join Anthropic, according to Business Insider. Jumper spent nearly a decade at DeepMind and shared the 2024 Nobel Prize in Chemistry with DeepMind CEO Demis Hassabis for AlphaFold’s breakthrough in predicting protein structures.
Two Departures in Three Days
Jumper’s exit is the second high-profile departure from Google’s AI research division this week. On June 18, Noam Shazeer, VP of Engineering at Google DeepMind and co-lead of Gemini, announced he was leaving Google to join OpenAI ahead of OpenAI’s confidential IPO filing. Shazeer, a co-author of the original Transformer paper, represented Google’s infrastructure lineage. Jumper represents something different: scientific reasoning applied to biological systems.
What Jumper Built
AlphaFold 3, the latest iteration of the protein-folding model Jumper co-developed, can predict the 3D structure of proteins from amino acid sequences with accuracy that matches experimental methods. The system’s architecture combines attention mechanisms with physics-informed constraints. It is one of a small number of AI systems that has directly accelerated scientific research, with applications in drug discovery, materials science, and molecular biology.
The Talent Signal
The two departures point in different directions for Google’s competitors. OpenAI recruited an infrastructure architect. Anthropic recruited a domain scientist. That distinction matters for what each lab is building next. Jumper’s expertise in structured prediction, physical constraints, and scientific reasoning aligns with Anthropic’s stated interest in agents that can operate in complex, domain-specific environments rather than purely general-purpose conversational systems.
What Google Loses
Google DeepMind now faces a concentrated talent drain at the VP level. Losing Shazeer weakened the Gemini foundation model team. Losing Jumper weakens the scientific applications portfolio that gave DeepMind its original identity as a research lab rather than a product organization. Whether Hassabis can retain remaining senior researchers, or whether the departures signal a structural pull toward labs offering equity, mission clarity, or faster deployment cycles, will shape Google’s AI competitiveness through the rest of 2026.