A mathematician at Anthropic says he has done what no one has managed in 167 years — proved the Riemann hypothesis, with the help of the company’s Claude chatbot. The Claude Riemann hypothesis proof claim, announced on Friday, has stunned the mathematics world. But the experts urging caution are just as loud as the ones calling it historic, because the proof has not yet been independently verified.
Levent Alpöge, a mathematician who works at Anthropic, published a four-page note on his personal website on Friday, November 20. In it, he wrote: “I believe I have, working with Claude, discovered a proof of the Riemann hypothesis.” He said the method also resolves the broader version of the problem, known as the generalised Riemann hypothesis.
He consulted Anthropic chief executive Dario Amodei before making the announcement. Anthropic itself declined to comment.
What the Claude Riemann Hypothesis Proof Claim Actually Says
The claim is, at this stage, exactly that — a claim. The proof has not been peer-reviewed, and no outside expert has yet confirmed it in full. The Clay Mathematics Institute, which offers a $1 million prize for solving the problem, still lists it as unsolved. Under the institute’s rules, a solution must be published in a peer-reviewed journal and then withstand two years of scrutiny before any prize is paid.
What Alpöge released is not a conventional written proof. He produced the argument in Lean, a programming language that allows mathematical arguments to be checked by a computer, and released only that version for now. A full written argument, he said, would be extremely long, and he acknowledged that he had not yet mastered all of the mathematics involved. He plans to write it up properly.
That admission is striking. “Claude far outpaced me near the finish line,” Alpöge wrote — an unusually frank confession from a mathematician announcing what could be the biggest result of the century.
Independent researchers have begun running the code. Andrew Sutherland of MIT said he verified it on his computer and had seen at least three other verifications. By Saturday, researchers had processed more than 175,000 Lean modules, with the computer’s kernel checks passing so far. But that is only part of the story: researchers were still testing whether the code’s definitions actually match the hypothesis as mathematicians understand it, and whether the argument holds together as mathematics, not just as code.
Why Mathematicians Are Not Celebrating Yet
The reaction from the mathematical community has been a mix of excitement and restraint — sometimes from the same person.
Brian Conrey, one of the world’s leading specialists on the Riemann hypothesis, called the announcement “a monumental moment in human history” and “the biggest thing I think that’s ever happened in mathematics.” He also admitted he had not yet seen a manuscript, only the remarks Alpöge posted on his website, and had asked for a copy. “It looks like it’s seriously complicated. I don’t think any human was about to find this,” Conrey said, adding that he believed “there will be simpler ways to do it.”
Others were blunter. Number theorist Andrew Granville of the University of Montreal called the released document “garbled” and said Alpöge should have prepared a clearer write-up before going public.
That tension is the heart of the story. A computer-checked proof is powerful evidence, but mathematics is a human discipline: a proof must be understood, not merely executed. Until other mathematicians can read the argument, follow its logic, and confirm that it proves what it claims to prove, the Riemann hypothesis remains open — whatever the computer says.
Nestor Guillen, a visiting professor of mathematics at New York University, welcomed Alpöge’s honesty about not fully understanding the proof, saying the back-and-forth he described showed AI working “more like a tool” and helped dispel “this idea of the machine god.”
What Is the Riemann Hypothesis — and Why $1 Million Hangs on It
First posed by the German mathematician Bernhard Riemann in 1859, the hypothesis concerns the distribution of prime numbers — the numbers divisible only by one and themselves. It says that every non-trivial zero of the Riemann zeta function lies on a single vertical line, where the real part equals one half. In plain terms, it claims a hidden order in the way primes are scattered through the number line.
It is widely regarded as the most important open question in pure mathematics. In 2000, the Clay Mathematics Institute named it one of seven Millennium Prize Problems, each carrying a $1 million reward. Riemann himself was born 200 years ago this September, which makes the timing of the claim — if it holds — almost poetic.
The hypothesis matters far beyond pure mathematics. Much of modern number theory is built on the assumption that it is true, and its resolution would reshape the field. That is why a claim of this size is treated with such care: the stakes are not just a prize, but the foundations of an entire discipline.
The Timing Problem: Anthropic’s IPO Was Yesterday
The announcement arrived the day after Anthropic’s record stock market debut — a coincidence Alpöge himself addressed head-on. He called the timing “horrible” and wrote that “there is an obvious cynical view, to do with the timing, but it is false,” adding: “I just could not live if I waited even longer.”
The context makes the timing hard to ignore. Anthropic went public this week in what was reported as the biggest IPO in history, and the company has been at the centre of the AI boom. Our earlier report on the listing is here: Anthropic set for biggest IPO in history.
There is also recent history. In August, Anthropic said an unreleased version of Claude had improved a long-standing result related to the Riemann hypothesis — raising the lower bound for the fraction of zeta zeros known to lie on the critical line from 41.6% to 67.2% — while stressing clearly that it had not proved the hypothesis. Friday’s announcement is a very different kind of statement.
And the claim comes two months after rival OpenAI said an AI model had solved another Millennium Prize Problem, on the equations that govern how fluids move. Mathematicians say that proof appears correct but has yet to be fully vetted. Alpöge, who had worked on related fluid problems with New York University mathematician Tristan Buckmaster, was at the centre of a dispute over credit for that work. OpenAI denied wrongdoing.
What Happens Next
Verification will take time — that is the one thing everyone agrees on. Mathematicians will need to read the Lean code, check that its definitions match the intended hypothesis, and translate the argument into a form humans can follow and judge. That process could take weeks, months, or longer, and it could still end in disappointment: history is full of claimed proofs of the Riemann hypothesis that collapsed under scrutiny. Whatever the outcome, the Claude Riemann hypothesis proof claim has already changed the conversation about AI in mathematics.
If the proof survives, it would be a landmark — the second Millennium problem possibly solved with AI help in as many months, and a moment that redefines what machines can do in mathematics. If it does not, it will still have shown how far AI has come: close enough to produce an argument that the world’s best specialists take seriously, even while they hold their applause.
For now, the honest summary is the one Alpöge himself used. He believes he has found a proof, working with Claude. Belief is not verification — but for a problem that has resisted every mathematician since 1859, it is already extraordinary news.
Sources: Phys.org / AFP reporting on the announcement (https://phys.org/news/2026-11-anthropic-claude-math-greatest-riddle.html).