Google DeepMind is an AI research laboratory founded in 2010 and acquired by Google in 2014. The organization conducts fundamental research in machine learning, neuroscience-inspired AI systems, and computational biology, operating from offices in London as part of Alphabet's broader AI strategy.
The lab has produced systems demonstrating capabilities in constrained domains. AlphaGo achieved documented performance against professional Go players through reinforcement learning and tree search. AlphaFold applied deep learning to protein structure prediction, generating predictions across the known proteome and enabling downstream experimental validation in structural biology.
Research spans multiple technical domains including game-playing AI, protein folding, reinforcement learning, and neuroscience-inspired architectures. Work emphasizes long-horizon research objectives rather than near-term product cycles, with stated priorities around building AI systems responsibly and advancing scientific understanding.