Anthropic Mathematician Levent Alpöge Claims Riemann Hypothesis Proof With Claude
Anthropic researcher Levent Alpöge says an unreleased version of Claude has proven the 167-year-old Riemann hypothesis, though mathematicians say the proof needs rigorous verification.

By Clara Tan, Artificial Intelligence
4 min read
Anthropic mathematician Levent Alpöge has claimed that an unreleased research version of the company's AI model Claude has proven the Riemann hypothesis, the 167-year-old conjecture about the distribution of prime numbers that carries a $1 million Millennium Prize.
Alpöge, a junior fellow at Harvard's Society of Fellows and researcher at Anthropic, posted the claim on 20 November 2026 alongside a PDF document titled "Remark on the Riemann Hypothesis" hosted on his personal site alpo.ge. He wrote on X that "Claude cracked math's most famous problem."
What the claim says
The Riemann hypothesis, proposed by Bernhard Riemann in 1859, asserts that all non-trivial zeros of the Riemann zeta function lie on a specific vertical line in the complex plane. Its truth would impose a precise form of randomness on the distribution of prime numbers. The Clay Mathematics Institute offers a $1 million prize for a correct proof.
Alpöge's claimed proof was produced by an unreleased research version of Claude, according to his post. The PDF document runs to more than 150 pages and includes a formally verifiable proof in Lean 4, a theorem-proving language.
Anthropic's prior published work
In August 2026, Anthropic published a research article describing how a version of Claude had improved a longstanding lower bound for the fraction of zeta zeros satisfying the hypothesis, from 41.6% to 67.2%. That work was validated by two Anthropic mathematicians and examined by experts Brian Conrey and Dan Goldston. Anthropic explicitly stated at the time: "We don't expect that the techniques Claude used will lead to proving the Riemann hypothesis."
Expert reaction
Nature reported on 20 November that "scholars say that the claim to have derived a proof for a landmark problem in mathematics will need rigorous checking." Scientific American similarly noted the claim "will need to withstand the scrutiny of the mathematical community." No independent verification has been published as of 21 November 2026.
Why it matters
If verified, this would represent the first time an AI system has solved a Millennium Prize problem, marking a watershed moment for AI in mathematical research. The Riemann hypothesis underpins vast areas of number theory and cryptography; a proof would have profound implications across mathematics and computer science.
However, the history of the Riemann hypothesis is littered with claimed proofs that did not survive peer review. The mathematical community's standard of verification is exceptionally high for this problem. Alpöge's Lean 4 formalisation may accelerate checking, but the process will take time.
Anthropic has not issued a separate statement on the November claim beyond Alpöge's personal post. The company's August research publication remains the only official account of Claude's mathematical capabilities on this problem.
Sources
- 1Levent Alpöge personal site (claim and PDF)alpo.ge
- 2Anthropic research: Claude improves Riemann zeta lower bound (Aug 2026)anthropic.com
- 3Nature: Mathematician says he and AI solved Riemann hypothesis (20 Nov 2026)nature.com
- 4Scientific American: Anthropic mathematician claims to prove Riemann hypothesis (20 Nov 2026)scientificamerican.com