Anthropic mathematician says he solved one of the ‘Millennium Problems’ with Claude

Today, 13:13950

Mathematician and Anthropic employee Levent Alpöge has said (.pdf) that, working with Claude artificial intelligence models, he has found a proof of the Riemann hypothesis, one of the seven ‘Millennium Problems’. This was reported by Scientific American.

The work has been published in Lean, a language designed for writing down and verifying mathematical proofs formally. It must now be assessed by specialists.

The Riemann hypothesis is one of the best-known unsolved problems in mathematics. It was formulated by the German mathematician Bernhard Riemann in 1859. It concerns the distribution of prime numbers and states that all non-trivial zeros of the Riemann zeta function have a real part equal to 1/2. A proof of the hypothesis could help to gain a deeper understanding of the patterns in the distribution of prime numbers.

In 2000, the Clay Mathematics Institute included the problem in its list of seven ‘Millennium Problems’, offering a prize of one million dollars for the solution of each. So far, the solution of only one of these problems has been officially confirmed: the Poincaré conjecture. It was proved by the Russian mathematician Grigori Perelman in the early 2000s.

Alpöge’s announcement came against the backdrop of another high-profile claim from the world of artificial intelligence. In the autumn of 2026, OpenAI announced that its AI agents had solved the problem of the existence and smoothness of solutions to the Navier–Stokes equations, which is also on the list of Millennium Problems. That claim also still awaits assessment by the mathematical community.

Against this backdrop, the announcement by the Anthropic employee has attracted particular attention: if the proof presented withstands independent verification, it will be one of the most significant mathematical achievements of recent decades.

Disclaimer: this article was translated by artificial intelligence and may contain inaccuracies.