Why DeepMind Disbanded the Nobel-Winning AlphaFold Team—and What It Means for Hassabis

Reorganization of the AlphaFold team

On 29 July, the Financial Times reported that DeepMind dissolved the independent AlphaFold team. Researchers formerly working on AlphaFold were reassigned to projects such as Gemini, AI programming, genomics, enzyme design, and nuclear‑fusion research, while some moved to Google’s biotech subsidiary Isomorphic Labs.

Key personnel—including Nobel Chemistry laureate John Jumper, Jonas Adler, and Alexander Pritzel—left DeepMind for Anthropic after a brief stint on the Code Strike team, which was created to improve Google’s AI coding capabilities in response to competition from Anthropic and OpenAI.

AlphaFold’s scientific impact

AlphaFold, launched in 2018, predicts protein three‑dimensional structures using AI. Since the 2021 launch of the AlphaFold Protein Database, the resource has stored predictions for over 200 million protein structures and has been accessed by more than 3 million researchers in over 190 countries. Official metrics show citations in more than 35 000 papers and inclusion as a research method in over 200 000 additional papers, driving advances in drug discovery, cardiovascular research, and agricultural breeding.

The DeepMind blog (published about 18 months earlier) highlighted that AlphaFold had become a “global engine for scientific discovery,” spurred the creation of Isomorphic Labs, and led to follow‑up models such as AlphaFold 3, AlphaGenome, and AlphaProteo.

Latest release and strategic shift

Hours before the Financial Times story, DeepMind released AlphaFold v3.0.4. Project repository: https://github.com/google-deepmind/alphafold3

DeepMind stated that the company’s strategy has moved from solving single scientific problems to building a Gemini‑centered, general‑AI system that can accelerate discovery across multiple domains. Gemini is positioned as a foundational platform for scientific innovation rather than a tool for any single project.

Industry context

The reallocation of resources aligns with broader AI‑industry dynamics: OpenAI, Anthropic, and other firms are heavily investing in large language models and AI agents. Google’s shift toward universal AI capabilities reflects an effort to maintain competitive advantage in this environment.