Anthropic AI "solved" Another "problem of the millennium" Anthropic worker Levent Alpege claims that with the help of AI assistant Claude, he was able to prove the Riemann hypothesis. It describes patterns in the distribution of prime numbers among natural numbers. Alpege published the proof in Lean, a special programming language for mathematics. According to him, more than 250 thousand pages of text will be required. The researcher himself admits that he has not yet understood the data obtained. "I will make every effort to understand this evidence and the simplifications used in it, and make it accessible to others," he said. Experts have already begun to study the computer code published by Alpege, also using AI, the dpa agency notes. The scientific community is still unable to confirm the correctness of Claude's "solution". The German mathematician Bernard Reimann formulated his hypothesis in 1859. Many of the mathematical statements of our time have been proven to "prove Riemann's hypothesis to be true", its cancellation would require the revision of all layers of science. In 2000, the American Clay Mathematical Institute included the Riemann Conjecture in its list of seven "things of the millennium" and awarded a million dollar prize for its proof. This is not the first time that AI claims to solve mathematical problems on a global scale. In the fall of 2026, OpenAI announced that its artificial intelligence agents had solved another of the seven "Millennium Challenges" - the problem of existence and smoothness of solutions to the Navier-Stokes equations. We are talking about basic equations that describe the movement of liquids and gases, from the flow of water in pipes to atmospheric processes. However, this decision has yet to receive official recognition from the mathematical community. @wglavnoe.