OpenAI Reports AI Models Made Breakthroughs in Math, Computer Science and Physics

OpenAI announced on October 27, 2026 that its AI models achieved three major breakthroughs in mathematics, computer science and physics, demonstrating an ability to independently discover and prove complex scientific theories rather than merely answering questions or assisting research.

In mathematics, the AI proved the local smoothing conjecture in all dimensions, which studies how long energy concentrations such as waves can persist, and thereby resolved the Kakeya conjecture and the Stein restriction conjecture. The Kakeya conjecture concerns the minimum space needed for a needle to point in every direction, while the Stein restriction conjecture relates to mathematical problems of complex wave behavior.

In computer science, the AI overturned the long-standing exponential time hypothesis, which holds that certain hard computational problems inevitably require rapidly growing computation as input size increases. OpenAI said the result could lead to faster algorithms in the future.

In physics, the AI proposed rigorous mathematical existence proofs for quantum chromodynamics, which explains how quarks combine into protons and neutrons, and for Yang-Mills theory, which describes forces between elementary particles and is tied to one of the seven Millennium Prize Problems.

OpenAI said the mathematics and computer science results have been formally verified using the Lean proof assistant and the proof materials have been made public. The physics result is currently under review by experts including a Nobel laureate in physics. The findings still require further independent confirmation from the academic community.