Google DeepMind has shut down the dedicated team behind AlphaFold, the Nobel Prize-winning artificial intelligence system that revolutionized protein structure prediction. Instead of maintaining specialized teams for individual scientific challenges, the company is redirecting researchers toward Gemini, its large language model (LLM), and other general-purpose AI initiatives. The AlphaFold system and its database of over 200 million predicted protein structures remain available to researchers worldwide, but the team that built it has been dissolved.

What Happened to the AlphaFold Researchers?

The shift represents a significant reorganization at DeepMind. According to reporting from the Financial Times, the majority of original AlphaFold paper authors were relocated to different departments starting in 2025. Rather than disappearing entirely, these scientists have been reassigned across multiple initiatives.

• Gemini Projects: Many former AlphaFold researchers have been moved to work on assignments related to Google's Gemini LLM, the company's flagship general-purpose AI system.
• Specialized Research: Some researchers have been assigned to work on genomics, nuclear fusion, and enzyme design, applying their expertise to new domains.
• External Moves: Approximately 25 percent of the full-time Google DeepMind authors of the original AlphaFold papers have left the company entirely, with some joining Alphabet's AI drug discovery firm, Isomorphic Labs.

This dispersion of talent reflects a broader strategic pivot at DeepMind. The company is moving away from building dedicated teams around specific scientific breakthroughs and toward a model where general-purpose AI systems support research across multiple domains.

Why Is Google Making This Change?

DeepMind's leadership has been explicit about the reasoning behind this reorganization. The company believes that general-purpose AI systems can tackle multiple scientific challenges more efficiently than specialized teams focused on single problems. This approach allows researchers to apply their skills more broadly across the organization's portfolio.

"Our strategy over the last nine years has been to focus on grand challenges... a concrete goal every project is focused on," said Pushmeet Kohli, Vice President of Research at Google DeepMind.

Pushmeet Kohli, Vice President of Research at Google DeepMind

The shift toward Gemini-backed systems suggests that Google believes large language models can be adapted to support scientific research more flexibly than purpose-built tools. Rather than creating new specialized teams for each breakthrough, the company is betting that general-purpose AI can be fine-tuned and applied to different scientific domains as needed.

What Does This Mean for AlphaFold's Future?

Despite the team's dissolution, AlphaFold itself is not disappearing. The AI system remains available to researchers worldwide, and the massive protein structure database it generated continues to support drug discovery and biological research. However, there will no longer be a dedicated Google DeepMind team maintaining and advancing the technology.

AlphaFold debuted in 2018 and solved a decades-old problem in structural biology: predicting protein structures with remarkable accuracy. The system has generated predictions for more than 200 million protein structures, providing an invaluable resource for scientists working on drug discovery and understanding disease mechanisms. Even without an active development team, this resource will continue to benefit the research community.

How to Stay Updated on AlphaFold and DeepMind's AI Research

• Access the Database: Researchers can continue accessing AlphaFold's protein structure predictions through public databases, which remain freely available for academic and commercial use.
• Monitor Gemini Developments: Follow Google DeepMind's announcements about Gemini-backed research initiatives to see how general-purpose AI is being applied to scientific challenges.
• Track Isomorphic Labs: Keep an eye on Alphabet's AI drug discovery subsidiary, where some former AlphaFold researchers have moved, to see how the technology is being commercialized.

The dissolution of AlphaFold's dedicated team marks a turning point in how major AI labs approach scientific research. Rather than building specialized teams around individual breakthroughs, Google DeepMind is consolidating its workforce around general-purpose AI systems. Whether this strategy will produce the same level of innovation as focused teams remains to be seen, but the company is clearly betting that Gemini and similar systems can deliver scientific breakthroughs across multiple domains simultaneously.