Co-founder of Google DeepMind, the AI laboratory owned by Alphabet. Hassabis founded DeepMind in 2010 with Shane Legg and Mustafa Suleyman. He holds a neuroscience PhD, which shaped his belief that the human mind should serve as a guide for building artificial general intelligence (AGI). Legg helped popularise the term AGI but did not coin it, contrary to some accounts.
DeepMind's early approach centred on reinforcement learning: teaching AI to play games, on the premise that gameplay requires problem-solving, as does science. The strategy took the lab from the arcade game Pong to the board game Go, in which its system defeated the world's top-ranked player, thrusting DeepMind into the global spotlight. DeepMind was a British company until Google bought it in 2014.
Hassabis attributes much of his success to making a theme-park game as a teenager in north London, and later founding a games studio. In April 2026 Google DeepMind released a paper on safeguards to prevent the development of AGI leading to disaster, co-authored by Legg.
In late January 2026 he said he would support a slowing of innovation to allow society to adapt to AI's social consequences.
For their work on AlphaFold, a program that can predict three-dimensional protein structures, Hassabis and John Jumper of DeepMind won half of the 2024 Nobel prize for chemistry. The protein-folding breakthrough, one of the hardest problems in biology, unlocked a universe of potential drugs.
Roughly five years ago Google spun off Isomorphic Labs, an AI-powered pharmaceutical firm, with Hassabis as its boss. The company's remit is to use DeepMind's AlphaFold technology to find novel medicines. As of April 2026 Isomorphic has 19 programmes across three main areas—cancers, cardiovascular conditions and immunology—including partnerships with Eli Lilly, Novartis and Johnson & Johnson. AlphaFold itself took six years to predict its first protein structure, and then one year to predict the structures of what Hassabis describes as "all 200m proteins known to science". He hopes a similar speedup will happen inside Isomorphic.
Pushmeet Kohli heads AI for Science work at Google DeepMind. Among the lab's tools are AlphaEvolve, designed to generate proofs for mathematical optimisation problems using natural language, and DeepThink AI, which can explain its reasoning to mathematicians. Terence Tao helped develop AlphaEvolve.
Air Force Inertia Axiom: Consistency is always easier to defend than correctness.