Story
Nov 20, 2026
Claude may have helped crack math’s biggest mystery—but the proof is still unverified
Anthropic mathematician Levent Alpöge says Claude helped him prove the Riemann hypothesis, a claim that could reshape mathematics if it survives scrutiny. Experts say the enormous proof needs time and independent review.
The Riemann hypothesis, a conjecture about the distribution of prime numbers, has resisted proof for 167 years. It is one of the Millennium Prize Problems, carrying a $1 million award for a verified solution. 1
Anthropic mathematician Levent Alpöge says he solved it on Sunday with help from the company’s Claude AI model. He described steering the system toward a promising mathematical setup, then watching it take the lead: “Claude flew, and danced, and sparkled near that finish line.” 1
Alpöge did not announce the result immediately. He said he first spoke with Anthropic CEO Dario Amodei and consulted other mathematicians, explaining that he did not want to release “slop” before the work was ready for scrutiny. His proposed proof is more than 250,000 pages long, and he has acknowledged that he does not yet fully understand all of its mathematics. 1
The claim became public on Friday, as reports cast it as a striking example of AI’s potential to unsettle mathematics. But the reaction from experts was measured. Cornell professor Steven Strogatz said the result would be “an extremely big deal” if it holds up, while warning that computer-based verification can make mistakes. “It’s fair to say that the claimed proof has not yet been independently verified,” he said. 1
Caltech mathematician Sergei Gukov also called the claim a “HUGE deal,” but UCLA scholar Terence Tao cautioned that digesting the proposed argument “will take some time.” Alpöge says he plans to help colleagues examine it. For now, the potential breakthrough remains a claim awaiting the mathematical community’s verdict. 1