Levent Alpöge, a mathematician at Anthropic, said Friday, Nov. 20, that Claude helped him crack the Riemann hypothesis, a problem that has resisted mathematicians for 167 years.
Alpöge published a four-page note and a 14-part Lean-formalized proof archive. He said he messaged Anthropic chief executive Dario Amodei late on Sunday, Nov. 15, about the finding. The announcement came a day after Anthropic's shares began trading on Nasdaq. Alpöge admitted: “The precise timing is horrible.”
The proof has not been independently verified by the mathematical community. The Riemann hypothesis was posed in 1859. The Clay Mathematics Institute offers a $1 million prize for a solution, but requires at least two years of public scrutiny before any award.
The same day, Stanford postdoctoral researcher Alessandro Masullo said he had proved the Generalized Riemann Hypothesis after 12 years of work without AI assistance. Stanford professor Ravi Vakil said of Masullo's work: “It is all human, with no AI.”
Masullo told fellow mathematicians: “We are still here. We are very much alive. We are not going anywhere.”
Alpöge's claim follows a string of AI-assisted math announcements this fall. OpenAI said in September it had solved the Navier-Stokes equations, and in October released 722 manuscripts grouped into 372 result clusters, 162 of them verified in Lean.
Toronto mathematician Daniel Litt offered a wary response: “There's just no more math in 50 years.”


