Levent Alpöge published a Lean proof on his website; independent verification and the Clay Institute’s review remain pending.
Levent Alpöge, an Anthropic mathematician, says he worked with Claude to produce a proof of the Riemann hypothesis and a formal Lean version without assumptions. He published the material on his website, and independent mathematicians have not yet verified it. Anthropic has not commented or involved itself in the announcement.
The Riemann hypothesis dates to 1859 and ranks among seven Millennium Prize problems that the Clay Mathematics Institute oversees, with a $1 million bounty for each solved problem. The hypothesis concerns the distribution of prime numbers and the zeroes of a function. Anthropic said in August that it was working on the problem with Claude but did not expect the techniques to yield a proof.
For builders and operators, the claim offers a test case for AI-assisted formal proof and for Lean-based verification pipelines. A verified result would give teams a strong benchmark for tooling that checks mathematical arguments, while an unverified claim highlights the need to separate model-generated drafts from accepted results. Teams should track independent review before treating AI math output as reliable.
Alpöge has published the Lean proof only on his website, and the Clay Mathematics Institute has not accepted it. Institute rules require publication and at least two years of scrutiny before any prize consideration. Alpöge also says he texted Anthropic CEO Dario Amodei late Sunday and suggests he delayed publication until after a Thursday IPO. Independent mathematicians now need to check each step.
What matters
- Levent Alpöge says he and Claude developed a proof of the Riemann hypothesis in Lean.
- Verification would give AI math tools a credible benchmark for proof generation.
- Independent reviewers and the Clay Mathematics Institute must scrutinize the proof for at least two years.
Why it matters
Independent reviewers and the Clay Mathematics Institute must scrutinize the proof for at least two years.
This GenAI News article was prepared in original wording using reporting and materials published by The Verge. Source reference: https://www.theverge.com/ai-artificial-intelligence/1010051/anthropic-riemann-hypothesis-millennium-prize-problem.
Drafted by the GenAI News review pipeline.
