Claude Claimed to Crack Riemann Hypothesis
An unverified Anthropic claim would make it the second Millennium Prize problem solved with AI, after OpenAI's reported work.
Updated on Nov. 21, 2026 in Artificial Intelligence

Live Poll
Do you think AI will be a net positive for society?
Anthropic mathematician Levent Alpöge claims that he and Claude, Anthropic's AI model, produced a proof of the Riemann hypothesis. The claim is unverified and has not been peer reviewed; Alpöge says the proof is written in the Lean programming language and is extremely long.
Why it matters
If verified, the proof would be the second Millennium Prize Problem solved with AI, following OpenAI's reported solution of the Navier-Stokes problem in September. The claim arrives as Anthropic launches its initial public offering, which puts a research assertion inside a corporate milestone.
The claim rests on about 250,000 pages of Claude-generated work, presented in Lean, a language that lets a computer check each logical step of a proof. The reported output is far larger than typical human proofs, but the stub gives no verification result, so the claim remains unconfirmed.
The players
Anthropic
An AI company whose Claude models are used for mathematical research, now going public with an initial public offering.
Clay Mathematics Institute
The nonprofit that established the Millennium Prize, which carries a $1 million award for each of seven problems, including the Riemann hypothesis.
The details
The Riemann hypothesis, proposed by Bernhard Riemann in 1859, concerns where the zeros of the Riemann zeta function lie, and it bears on the distribution of prime numbers. Alpöge wrote on his website that he believes he discovered a proof working with Claude and used Lean, a computer programming language that checks proofs step by step, to present it. He described the argument as incredibly long and said Claude produced about 250,000 pages to reach the solution. The problem's status as a Millennium Prize Problem means a verified proof would carry a $1 million award from the Clay Mathematics Institute.
Timeline
1859: Bernhard Riemann proposes the Riemann hypothesis.
2000: The Clay Mathematics Institute's Millennium Prize lists the problem among seven $1 million challenges.
September: OpenAI reportedly solves the Navier-Stokes problem.
Sunday night, hours before the IPO: Alpöge reportedly discovers the proof.
November 20, 2026: Anthropic launches its initial public offering.
The Tech Race
OpenAI's reported solution of the Navier-Stokes problem in September set a precedent for AI-assisted work on Millennium Prize problems. The Riemann claim places Anthropic in that same contest, though neither result has been independently confirmed in the stub.
The claim does not change anything for general users today, since no product is involved beyond Claude's use in research. Readers should treat the result as unverified until mathematicians have checked the proof.
The takeaway
Watch for independent mathematical review of the Lean proof and any public response from the Clay Mathematics Institute. Verification or rejection of the argument would settle whether AI has solved a second Millennium Prize problem.
Further reading
For broader context on AI tools in research, see the Artificial Intelligence section.
Live Poll
Do you think AI will be a net positive for society?



