Anthropic Chief Commercial Officer Paul Smith, center left, and Anthropic Chief Financial Officer Krishna Rao, center right, alongside others, react during the Anthropic IPO closing ceremony at the Nasdaq MarketSite in New York, Thursday, November 19, 2026.
Seth Wenig/AP
hide caption
Seth Wenig/AP
A mathematician who works for artificial intelligence giant Anthropic claims to have solved a holy grail of math problems with the help of the company’s AI tool, Claude.
This is the second time in recent months that an AI company or someone linked to a company has claimed to have solved a famous unsolved math problem.
Levent Alpöge, a member of the company’s technical staff, made the announcement on Friday through a four-page explanation posted on its website. He said he believed that while working with Claude he had found proof of the Riemann hypothesis.
“Claude and I have had a great time together over the past few months, but this time Claude was way ahead of me near the finish line,” Alpöge wrote. “No, Claude flew, danced and sparkled near the finish line.”
Formulated in 1859, the Riemann hypothesis focuses on the distribution of prime numbers – those that are only divisible by one and themselves – as they grow toward infinity.
Alpöge said his discovery took place on Sunday evening. He messaged Anthropic CEO Dario Amodei to ask about “my ability to announce,” as the San Francisco-based company was about to make news with its stock market debut.
A day after the company’s IPO – the most valuable in history – Alpöge revealed it had solved the problem that had proved elusive for more than 150 years. Anthropic did not respond to an email from The Associated Press seeking comment on Alpöge’s statement.
Alpöge published his and Claude’s proof of the Riemann hypothesis in 17 million lines of Lean code, a programming language that allows users to program proofs of mathematical theorems. Traditionally, the results buried in this code would be published in a formal paper, said Alex Kontorovich, distinguished professor and chair of the mathematics department at Rutgers University.
“Usually when someone gives you an article, it’s also kind of code for proof and you have to actually read the article, follow it yourself, determine whether you believe every line of that article. And when you’re done, you should have a deep understanding of why something is true,” he said.
He said Claude followed this code and the evidence was true.
“Does that mean I understand? Not yet. But the computer is extraordinarily picky, way pickier than me and so if you can satisfy the computer, there’s a very, very good chance you’ll satisfy me too,” he said. “But I have to do the work.”
Hours after the announcement, the Riemann hypothesis remained listed as unsolved on the Clay Mathematics Institute website. The institute selected this hypothesis in 2000 as one of the seven Millennium Problems and offered a million-dollar reward for the solution.
The process of evaluating whether the problem has actually been resolved takes time, said Martin Bridson, president of the institute.
“This is an announcement of historic significance, one that promises a dramatic transformation in the perception of what is possible in the fight to expand human understanding of mathematics,” he said in a statement. “But the road from today’s announcement to real understanding will be long and difficult.”
Alpöge said that he “presents to the world a complete argument before he has mastered a vast neighborhood around his mathematics.” He said more than 250,000 pages came from his work with AI.
In September, OpenAI said its technology was key in a similar breakthrough with Navier-Stokes, an equation that governs fluid flow.
“Navier-Stokes fell and now the Riemann hypothesis,” Kontorovich said. “There’s a sort of out-of-body experience where I look at myself and say, ‘Yes, the Riemann hypothesis has been solved,’ and I’m sort of trying to understand what those words are coming out of my mouth.”
Anthropic made a blockbuster Wall Street debut this week, raising a record $81 billion at a valuation of $1.5 trillion.
Amodei projected an optimistic vision of the future of AI in which the technology leads to significant advances in mathematical and scientific research while warning of the potential risks that AI poses.
Gn headline

