Google has disbanded its Nobel Prize-winning team. Some of its members will join Anthropic

Google DeepMind, Google’s artificial intelligence research division, has disbanded the research team that created the AlphaFold AI system, which was capable of predicting the three-dimensional structure of proteins and accelerating drug development, the Financial Times has learned from sources familiar with the situation and an analysis of personnel changes.

A month ago, one of the project leaders—Google DeepMind Vice President John Jumper—left AlphaFold. In 2024, he and the Google project team were awarded the Nobel Prize for the system developed as part of AlphaFold to predict the three-dimensional structure of proteins.

Jumper's announcement of his departure sent Alphabet's stock tumbling 7%; in trading on July 29, the stock is up just over 1%.

Over the past year, most members of the AlphaFold team have been reassigned to new projects, the newspaper reported. DeepMind confirmed that researchers from AlphaFold are now working on projects based on the Gemini AI model, as well as on developments in the fields of enzymes, fusion energy, and genomics. Some of the specialists have moved to Isomorphic Labs—an Alphabet division focused on drug discovery using AI—while nearly a quarter of the scientists listed as authors of the original scientific papers on the AlphaFold system have left the company altogether.

One of the most notable departures from Google was John Jumper’s resignation last month. Earlier this year, he and AlphaFold researcher Jonas Adler were transferred to Google’s internal Code Strike team, which develops AI for programming, and in June 2026, he announced his move to Anthropic. Adler and another key member of the AlphaFold project—Alexander Pritzel—will also be joining him at Anthropic, the newspaper reported. According to a Financial Times source at DeepMind, their departure came as a surprise to the company’s employees.

Google’s AI lab has changed its approach to scientific research, DeepMind Vice President of Research Pushmit Kohli told the Financial Times. Whereas teams were previously organized around specific fundamental tasks, such as predicting protein structures, the priority now is to create Gemini-based systems capable of assisting scientists and, over time, automating certain stages of scientific work. Google is competing with OpenAI and Anthropic in the development of cutting-edge AI agents, the FT reports.

AlphaFold, whose development began in 2018, has become one of DeepMind’s most famous scientific achievements, the FT notes. The system has learned to predict the three-dimensional structure of proteins with high accuracy, which has significantly accelerated disease research and the development of new drugs. This breakthrough earned DeepMind CEO Demis Hassabis and lead researcher John Jumper the Nobel Prize in Chemistry in 2024.

The current reorganization at Google reflects a broader shift in the AI industry, notes the Financial Times. Major research labs are increasingly viewing large language models not only as chatbots, but also as versatile tools for accelerating scientific research. At the same time, competition for top talent is intensifying: companies are actively poaching scientists capable of creating cutting-edge AI systems.

This article was AI-translated and verified by a human editor