Anthropic researcher claims AI-assisted proof of Riemann Hypothesis
Anthropic mathematician Levent Alpöge says he used Claude over six weeks to generate a proof of the Riemann Hypothesis, a problem first posed by Bernhard Riemann in 1859 about the distribution of prime numbers. The proof, written in the verification language Lean, reportedly spans more than 250,000 pages and has not yet been formally confirmed. Oxford mathematician James Maynard told The New York Times he is '98 or 99 percent certain' the result is correct.
GoKawiil's interpretation of the reporting above, not reported fact.
The Riemann Hypothesis is one of the Clay Mathematics Institute's Millennium Prize Problems, carrying a $1 million prize, so independent verification will likely take considerable time given the proof's length. If confirmed, it would suggest AI systems like Claude can meaningfully assist in solving century-old open problems, following OpenAI's similar push into unsolved math problems in October. The claim also raises questions about how mathematicians and institutions will validate extremely long, AI-generated proofs.
- Anthropic's Levent Alpöge claims Claude helped produce a proof of the Riemann Hypothesis in six weeks.
- The proof, written in Lean, is reportedly over 250,000 pages and still needs verification.
- Oxford's James Maynard expressed high but not total confidence in the proof's accuracy.
Source: engadget.com — Mariella Moon, 2026-11-21
Published there as: “Anthropic employee claims to have proven the most famous math problem using Claude”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.