Anthropic mathematician Levent Alpöge announced on Nov. 20 that he and the company’s Claude AI had produced a formal proof of the Riemann hypothesis, a 167-year-old problem that carries a $1 million Millennium Prize.
- The proof is encoded in 17 million lines of Lean code and spans more than 250,000 pages of mathematical argument.
- Independent researchers have compiled the Lean code without errors, though a full human-readable explanation does not yet exist.
- The Clay Mathematics Institute still lists the hypothesis as unsolved; its rules require peer-reviewed publication and two years of scrutiny before awarding the prize.
- A separate, human-only proof was posted the same day by a Stanford postdoctoral researcher, Alessandro Masullo.
Alpöge, a member of Anthropic’s technical staff, posted a four-page note on his website Friday, Nov. 20, stating he had worked with a non-public version of Claude for about six weeks. “I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis,” he wrote, according to New Scientist. The proof was released directly in Lean, a programming language whose small kernel verifies every logical step.
If the Lean code compiles error-free, the mathematics community generally treats that as strong evidence of correctness. The hypothesis, first proposed by German mathematician Bernhard Riemann in 1859, concerns the distribution of prime numbers — those divisible only by themselves and one — and is widely regarded as the most important open question in pure mathematics. It conjectures that a hidden pattern underlies the appearance of primes along the number line.
The formalization is massive: 17 million lines of Lean code and a human-readable version that Alpöge said exceeds 250,000 pages. He acknowledged he had not yet mastered all of the mathematics involved and planned to write a proper exposition. “I am clearly unilaterally responsible for feeling I just could not live if I waited even longer.” “There is an obvious cynical view, to do with the timing, but it is false,” he wrote, referring to Anthropic’s initial public offering the day before — the largest in history, raising $81 billion at a $1.5 trillion valuation.
The New York Times's coverage (by ace reporters @kchangnyt and @sioroberts) of today's remarkable events regarding the Riemann hypothesis and the Navier-Stokes unforced blowup problem. https://t.co/nMkBnllrsU
He described the timing as “horrible” and said he messaged Anthropic CEO Dario Amodei on Sunday night to ask how to announce the finding. Alpöge also said the method resolves a broader version known as the generalized Riemann hypothesis.
Within hours, mathematicians on the Zulip chat server began testing the Lean code. Andrew Sutherland at MIT confirmed it compiled successfully. Kevin Buzzard of Imperial College London noted two remaining checks: whether the Lean statement actually matches the Riemann hypothesis, and whether the code compiles on independent kernels. Both checks have since passed. Buzzard wrote on the Lean community chat server that the kernel run has now been completed.

The Clay Mathematics Institute, which selected the Riemann hypothesis as one of seven Millennium Prize Problems in 2000, still lists it as unsolved on its website. Under the institute’s rules, a solution must appear in a peer-reviewed journal and withstand two years of scrutiny before the $1 million prize can be awarded. Alpöge has so far released only the Lean formalization, not a traditional journal paper.
In September, OpenAI said its technology helped solve a Millennium Prize Problem related to the Navier-Stokes equations, though that proof has not been fully vetted. OpenAI later released claimed solutions or progress on 372 additional problems. Guillen added he was glad that “this time there might not be any of the same corporate controversy.”
Simple explanation of the usefulness of the Riemann proof (so far) is that it removes the conditional nature of other math theorems dependent upon it. And, of course, is yet another mind-blowing thing achieved by super intelligence …https://t.co/F1sAyxwaSF
More on Levent Alpöge
Even if Riemann turns out not to be fully solved yet, it will be soon enough https://t.co/wqYQ6FI05l
If this proves to be true, it will be a monumental breakthrough. An ode to mathematics, to AI, and above all, to human ingenuity. After all, there is no such thing as AI born out of AI only.https://t.co/3NJmjnIJcD
Anthropic's Levent Alpöge just claimed a Riemann Hypothesis proof with Claude. Here's what you need to know. Alpöge, a mathematician at Anthropic, posted a paper called remark.pdf at https://t.co/gM0DfommLn. He writes, "I believe I have, working with Claude, discovered a proof… https://t.co/7V9N3HUuTr pic.twitter.com/GiprmBBI9t