TL;DR
Anthropic mathematician Levent Alpöge says he worked with Claude on proofs of the Riemann Hypothesis and the Generalized Riemann Hypothesis. He released a four-page note and a large Lean code archive, but the claim has not been accepted: a key code check remains pending, and mathematicians have not established that the argument is correct.
Anthropic mathematician Levent Alpöge said on November 20 that he believes he found proofs of the Riemann Hypothesis and the Generalized Riemann Hypothesis while working with Claude. He published a four-page note and a large archive of Lean code, but the claim remains unverified: mathematicians have not established that the code proves the stated results, and no journal or mathematical body has accepted the work.
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
Alpöge’s note, posted on his personal website, says he believes he discovered a proof of the Riemann Hypothesis with Claude. It also claims that the method extends to the Generalized Riemann Hypothesis, a broader statement concerning a family of related functions. These are claims made by Alpöge; the note is not a peer-reviewed paper and does not present a readable proof for experts to assess.
Instead, the note links to a formalization split across 14 parts of Lean code. Lean is software that checks mathematical arguments written in a formal language. Computer scientist Scott Aaronson put the archive at about 17.4 million lines of code. Alpöge says a human-readable version would exceed 250,000 pages and that he chose not to post it. That scale has made it difficult for mathematicians to inspect the argument directly.
There has been an initial report that the archive compiles independently. MIT mathematician Andrew Sutherland told Scientific American that it compiled “with only the three standard axioms.” However, mathematician Kevin Buzzard said a “comparator run” was still awaited. That check is intended to test whether the formal code establishes the claimed statement using the intended definitions and assumptions. Compilation alone does not settle whether the code proves the result Alpöge describes.

A Millennium Problem Under Review
The Riemann Hypothesis has remained open since Bernhard Riemann proposed it in 1859. It concerns the locations of the nontrivial zeros of the zeta function and is closely connected to how prime numbers are distributed. A correct proof would resolve one of mathematics’ best-known unsolved problems and meet one of the seven Millennium Prize Problems set by the Clay Mathematics Institute, which offers $1 million for an accepted solution.
The episode also draws attention to how AI systems might contribute to advanced mathematics. Alpöge says Claude did a substantial share of the technical work, while he guided the strategy. That account is significant if the proof withstands scrutiny, but it does not establish how much the model contributed or whether the result is correct. For now, the public evidence is an AI-assisted claim and a formal code archive under review.
Popular right now

Dyson V8 Cyclone Cordless Vacuum

KitchenAid Artisan 5-Qt Tilt-Head Stand Mixer, Almond Cream

Ninja Foodi 10-Quart DualZone 6-in-1 Air Fryer, Black, DZ302
As an affiliate, we earn on qualifying purchases.
From Riemann’s 1859 Conjecture
The Riemann Hypothesis concerns the nontrivial zeros of the Riemann zeta function. In simplified terms, it predicts that they all lie on a particular vertical line in the complex plane, where the real part is one-half. The conjecture is important partly because its truth would sharpen mathematicians’ understanding of the distribution of prime numbers. It has resisted proof for more than a century and a half.
The Generalized Riemann Hypothesis extends the question to related functions. Alpöge’s note says the method “easily extends much further,” but the note itself does not supply a readable argument establishing that extension. The two claims should therefore be treated as assertions in the posted note, not as results confirmed by the mathematical community.
Lean formalizations can help verify that each step in a proof follows within a specified system. They still require scrutiny of what has been formalized, which definitions and assumptions are used, and whether the checked theorem matches the claim being made. A successful compile is one part of that process, rather than a substitute for expert review and mathematical acceptance.
“I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis”
— Levent Alpöge, in his November 20 note
Key Checks Remain Open
The central question is whether the Lean archive proves the precise claims made in Alpöge’s note. The source material reports an independent compile but also says the comparator run was still pending as of November 21. It does not report a completed result from that test, a detailed independent mathematical review, or a human-readable proof that experts have followed from start to finish.
Other details are also unreported. Alpöge’s note does not identify which Claude model was used, and Anthropic had not announced the claim on its news page as of November 21. Nature reported that the company had not immediately responded to its request for comment. The source material also describes mathematician Andrew Granville as calling the released note “garbled,” underscoring that the short note has not resolved the questions raised by the archive.
It is therefore unclear whether the archive will pass the remaining checks, whether the claimed extension to the Generalized Riemann Hypothesis is supported, and how researchers would assess Claude’s role. Until those points are addressed, the proof claims remain unconfirmed.
Formal Review and Publication
The next immediate milestone is the comparator run identified by Buzzard, followed by scrutiny from mathematicians able to assess the definitions, assumptions and formalized statement. Researchers would also need to determine whether they can produce or follow an intelligible account of the proof’s mathematical ideas. The supplied sources do not give a timetable for those steps.
Any claim to the Clay prize faces a longer process. The institute’s rules require publication in a refereed journal of worldwide repute, at least two years of scrutiny after publication, and general acceptance by mathematicians. The prize process has not begun on the information available as of November 21. The outcome will depend on what independent checks and expert review establish.
Key Questions
Has the Riemann Hypothesis been proved?
Not on the evidence reported so far. Alpöge says he believes he and Claude found a proof, but mathematicians have not accepted the claim and a key code check was still pending as of November 21, 2026.
What did Alpöge publish?
He posted a four-page note linking to a 14-part archive of Lean code. The note gives a short sketch and timeline, rather than a readable proof of the claimed results.
What role did Claude play?
Alpöge says Claude helped with the work and “far outpaced” him near the finish. The note does not identify the Claude model, and the extent of the model’s contribution has not been independently established in the material available.
Does a Lean compile confirm the proof?
No. An independent compile was reported, but a comparator run was still awaited. Even a successful code check must be assessed to determine whether the formalized result matches the theorem being claimed.
Could this qualify for the Clay Mathematics Institute prize?
The institute offers $1 million for an accepted solution to the Riemann Hypothesis, under rules that include refereed journal publication, at least two years of scrutiny and general acceptance by mathematicians. The reported claim has not reached those stages.
Source: Thorsten Meyer AI
