The claim that stunned the mathematical world last week has passed its first major test. Two organizations at the center of the formal mathematics ecosystem announced on Saturday that they had independently verified Anthropic mathematician Levent Alpöge's proof of the Riemann hypothesis — the 167-year-old problem widely considered the most famous open question in mathematics — which Alpöge says he produced working roughly six weeks with Anthropic's Claude.

The announcement, reported by The Guardian, converts a dramatic claim into a verified result, at least in the eyes of the groups best positioned to judge it. If it continues to hold up under broader scrutiny, it would mark the first time an AI-assisted effort has resolved a problem long regarded as mathematics' most storied unsolved challenge — and a milestone in the rapidly intensifying contest between frontier AI labs to demonstrate machine-driven scientific discovery. For more context on this story, see our ongoing artificial intelligence updates.

The Verification: Lean FRO and Mathlib Initiative Sign Off

Alpöge released his argument not as a traditional paper but as computer-checkable code written in Lean, a programming language in which every logical step of a proof can be mechanically verified. On Saturday afternoon, the Lean Focused Research Organization — a group dedicated to advancing the use of the language — and the Mathlib Initiative, which maintains Lean's central mathematics library, announced that they had reviewed the formalization and confirmed the result.

"We have full confidence that Alpöge's proof of the Riemann hypothesis is correct, and we congratulate him on this achievement," the two groups said, according to The Guardian.

That statement carries unusual weight precisely because of how the proof was released. A complete Lean formalization that compiles against standard axioms leaves no room for the subtle gaps that have sunk many famous claimed proofs over the decades — the computer either accepts every step or it does not.

A Proof Released as Code, Not Paper

The story began on Thursday, when an unsigned four-page note titled "Remark" appeared on Alpöge's personal website. Its opening line was direct: "I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis."

The note itself contained only a sketch. The full argument, Alpöge wrote, had been entirely formalized in Lean, with the files hosted on the same site split into 14 compressed archive parts. No human-readable paper accompanied the release; Alpöge said the argument was too long to write up in polished form for now, promising a readable version later. A footnote in the note joked that the informal, AI-generated draft material behind the proof ran past 250,000 pages.

The note also claimed more than the classical result: according to coverage by OfficeChai, it says the same method proves the Generalized Riemann Hypothesis, which covers a broad family of related functions, with a further extension expected "with more time." A 20-step timeline, which the note says Claude compiled, runs from early October to mid-November 2026, with the final step dated November 16.

"Claude and I have had a great time together with such breadth in recent months, but this time Claude far outpaced me near the finish line," Alpöge wrote. "No, Claude flew, and danced, and sparkled near that finish line."

Why the Riemann Hypothesis Matters

Proposed in 1859 by the German mathematician Bernhard Riemann, the hypothesis concerns the zeros of the Riemann zeta function, a mathematical object that encodes how prime numbers are distributed along the number line. Riemann conjectured that all of the function's "nontrivial" zeros lie on a single vertical line in the complex plane. If true, it implies the primes are scattered about as regularly as they possibly could be — a statement with deep consequences across number theory and cryptography.

The claim has been checked computationally for the first 10 trillion solutions, according to the Clay Mathematics Institute, but verification of examples is not proof. The Riemann hypothesis is one of seven Millennium Prize Problems selected by the institute, which offers $1 million for a solution to each. It has resisted the efforts of the field's greatest minds for over a century and a half.

There is a fitting symmetry in the timing: 2026 marks Riemann's bicentenary. The mathematician who named the problem was born 200 years ago this September, according to Nature.

'Historic Milestone' — and Hard Questions About Timing

Reactions from leading mathematicians have blended awe with pointed skepticism. "If this fully checks out, it is clearly a historic milestone in mathematics," James Maynard, a Fields Medal-winning number theorist at the University of Oxford, told Nature.

Others focused on how the announcement was handled. Andrew Granville, a number theorist at the University of Montreal, criticized the four-page write-up as insufficient. "What he released is garbled," Granville told Nature, arguing Alpöge should have prepared a clearer document before going public.

The timing also raised eyebrows. The note appeared one day after Anthropic's initial public offering on the Nasdaq — an IPO that initially valued the company at $1.5 trillion before its stock dropped on November 19, according to Nature. "The precise timing is horrible. If I could have, I'd have finished this all much earlier," Alpöge wrote, adding in the note itself: "There is an obvious cynical view, to do with the timing, but it is false."

The Human Counterpoint

The AI-assisted claim was not the only dramatic development of the week. Hours after Alpöge's note appeared, Ravi Vakil, a mathematician at Stanford University, posted a rival preprint in what he called a "hurried announcement," according to Nature. The rival work was written by Alessandro Maria Masullo, who earned his PhD at Stanford in 2020 under number theorist Brian Conrad.

The New York Times reported that Masullo, an independent mathematician, said he had worked on the problem for nearly 12 years — and that he deliberately did not use AI, explaining that he did not fully trust the systems. The dueling manuscripts set off what observers described as the largest mathematics discussion the internet has seen this year.

What Comes Next

Saturday's verification is a landmark, but several questions remain open. The promised human-readable write-up of the proof has not yet been published, leaving many mathematicians unable to study the argument in its traditional form. Wider community scrutiny — the process by which a proof becomes accepted mathematics — is only beginning.

The episode also lands amid growing unease about the role of private AI companies in fundamental research. In September, OpenAI claimed that a non-public version of its tools had helped resolve part of a different Millennium Prize Problem, in a project The Guardian reported cost $15 million and employed 10,000 AI systems. University of Warwick mathematician James Robinson told the Guardian at the time that it was "frustrating to see these big tech companies burning this fuel up just so that they can show off about how great their latest model is."

Whether the Riemann hypothesis chapter closes with Alpöge's verification or opens into a longer debate about AI's place in mathematics, one thing is already clear: the way discoveries are made — and announced — in the field may never look quite the same.

---

Stay Ahead of AI

Get the latest AI news, analysis, and breakthroughs — all in one place.

Read more AI news →