Story
Nov 22, 2026

A rival human proof turns the Riemann claim into an AI showdown

An Anthropic mathematician says Claude helped him prove mathematics’ most famous open problem, but a Stanford researcher has now posted a competing proof developed without AI. Experts have yet to verify either result.

On Friday, Anthropic mathematician Levent Alpöge said he had completed a proof of the 167-year-old Riemann hypothesis with help from Claude. Rather than publish a conventional paper, he posted a short note linking to computer code he said could verify the result. “Claude flew, and danced, and sparkled near that finish line,” Alpöge wrote.

The announcement drew immediate excitement from AI optimists. Google CEO Sundar Pichai celebrated: “Somewhere Euler, Gauss and Riemann in awe today!” But mathematicians urged caution. Cornell professor Steven Strogatz said Alpöge had played a central role in guiding Claude, likening the collaboration to “a very skillful jockey riding a really fast horse.” He also warned that computer-based checking can make mistakes; the proof had not been independently verified.

Hours later, Stanford postdoctoral researcher Alessandro Masullo announced a rival proof, which he said he had developed without AI over 12 years. He said Alpöge’s claim had caught him about a year from releasing his own work, and that he was posting a version that might contain errors because it had been rushed. Masullo cast his announcement as a defense of human mathematicians: “We are still here. We are very much alive. We are not going anywhere.”

The competing claims have made a landmark problem into a test of both AI’s mathematical ability and human expertise. Researchers are now scrutinizing both proofs, and neither had been formally established. Elon Musk hailed the supposed breakthrough as “yet another mind-blowing thing achieved by super intelligence,” while adding that even if the hypothesis was not fully solved, it would be soon. The prize—and the verdict—remain out of reach.

Story coverage