Story
Nov 21, 2026

Two mathematicians claim the Riemann hypothesis is solved—one with AI, one without

Levent Alpöge says Claude helped him prove the 167-year-old Riemann hypothesis; Alessandro Masullo says he reached a proof on his own. Neither claim has been formally verified, leaving mathematicians to assess two rival solutions.

Levent Alpöge, a mathematician on Anthropic’s technical staff, said Friday that he had proved the Riemann hypothesis with help from Claude. He said he reached the result on Sunday, then held off announcing it while consulting other mathematicians. “Claude flew, and danced, and sparkled near that finish line,” he wrote. The claim has not been independently verified; Alpöge described a proof more than 250,000 pages long, and Cornell mathematician Steven Strogatz cautioned that a computer-based check can make mistakes.

Hours later, Alessandro Masullo, a Stanford postdoctoral researcher, said he had a proof of his own—developed without AI over 12 years. He said Alpöge’s announcement caught him about a year before he planned to release his results, prompting him to publish a version he warned might contain errors. “We are still here,” Masullo wrote, arguing that human mathematicians remain part of the story.

The competing claims concern a problem first proposed 167 years ago, involving the distribution of prime numbers. It is one of the Millennium Prize Problems, with a $1 million award for a solution. Neither proof had been formally established as correct by Friday afternoon, and mathematicians were working to scrutinize both.

The news also drew reactions from tech leaders. Google CEO Sundar Pichai posted, “Somewhere Euler, Gauss and Riemann in awe today!” Elon Musk struck a more cautious note: “Even if Riemann turns out not to be fully solved yet, it will be soon enough.”

The central question remains unresolved: whether either argument holds up—and what that would mean for mathematics in an era of increasingly capable AI.

Story coverage