🔍 Read the full analysis: Can AI Assist With A Riemann Hypothesis Proof? The Claude Claim on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Levent Alpöge, a mathematician who works at Anthropic, posted a claim on November 20 that he and Claude found proofs of the Riemann Hypothesis and the stronger Generalized Riemann Hypothesis. The released Lean code has passed some initial checks, but a key comparison test was still pending, mathematicians have not established that the argument proves the stated results, and no proof has been accepted.
Mathematician Levent Alpöge, who works at Anthropic, said in a note posted November 20 that he believes he and Claude have proved the Riemann Hypothesis, a famous unsolved problem about the distribution of prime numbers. The note links to a large Lean proof archive, but mathematicians have not accepted the result, and a key check of whether the code proves the intended statement was still pending as of November 21.
Alpöge’s four-page note says the method also proves the Generalized Riemann Hypothesis, a broader family of related conjectures. It offers a short sketch and a timeline rather than a conventional proof, and links to a 14-part archive of Lean code. Lean is software that checks whether the steps of a formal mathematical argument follow from its stated assumptions. Computer scientist Scott Aaronson estimated the archive at 17.4 million lines of Lean code, according to the source material.
Alpöge says the result is too large to present in a readable form and says he did not publish the more than 250,000-page version he generated. He describes himself as setting the strategy while Claude handled much of the technical work, writing that “Claude far outpaced me near the finish line.” The note does not identify which Claude model was used. These descriptions of the collaboration come from Alpöge; there is no independent account of how the work was divided.
There have been some preliminary technical checks, but they do not settle the mathematical claim. MIT mathematician Andrew Sutherland told Scientific American that the code compiled independently “with only the three standard axioms.” Kevin Buzzard wrote that a “comparator run” was still awaited. That test is intended to check whether the formalized result matches the conjecture being claimed, rather than a different or weaker statement. Compilation alone does not establish that the code proves the Riemann Hypothesis.
Explainer · Riemann hypothesis claim
From claim to accepted proof: six checks, and where this one stands
What the hypothesis says, and the path every proof claim has to take before mathematicians accept it.
The claim in one picture
Riemann’s guess (1859): every non-obvious zero of the zeta function lies on this one line. If true, the primes are spread out as evenly as possible.
The checks a proof claim must pass
- 1Claim postedA 4-page note plus Lean code, November 20, 2026✓ done
- 2Code compiles elsewhereIndependent compile reported by an MIT mathematician✓ reported
- 3Comparator runRight statement? Only standard axioms? Several Lean kernels?… pendingWe are here · November 21, 2026
- 4Humans understand itA readable proof that experts can follow○ open
- 5Journal publicationA refereed mathematics journal of worldwide repute○ open
- 6Two years + general acceptanceWhat the Clay prize rules require before a $1 million award○ open
Sources: Clay Mathematics Institute (Riemann Hypothesis page; Millennium Prize rules); Levent Alpöge, remark.pdf (November 20, 2026); Scientific American, November 20, 2026 (independent compile; comparator run awaited); Kevin Buzzard, Xena Project, November 12, 2026 (the checks). Status as reported in these sources by November 21, 2026.
Graphic: Powered by Thorsten Meyer AI
A Claim With Major Mathematical Stakes
If independently verified, a proof would resolve a problem that has shaped number theory since Bernhard Riemann formulated the conjecture in 1859. The Clay Mathematics Institute lists it among its Millennium Prize Problems, with a $1 million prize attached under the institute’s rules. The hypothesis concerns the non-obvious zeros of the zeta function; its truth would give mathematicians information about how prime numbers are distributed.
The claim also concerns AI-assisted mathematics. Alpöge says Claude contributed substantially to a result whose formal proof is too large for humans to review line by line in its current form. The work raises questions about how mathematicians can verify machine-generated arguments, whether formal proof software can make such work auditable, and how credit should be assigned when a person and an AI system work together. These questions are separate from whether this proof is correct.
James Maynard, a Fields Medal winner, told Nature, “If this fully checks out, it is clearly a historic milestone in mathematics.” Maynard’s assessment is conditional on verification and expert scrutiny, neither of which is complete.
Products Worth Considering

FLEXISPOT 36-Inch Height-Adjustable Standing Desk Converter with Keyboard Tray

Vari ComfortEdge 48×30 Inch Adjustable Electric Standing Desk

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
As an affiliate, we earn on qualifying purchases.
Why the Riemann Claim Matters
Prime numbers such as 2, 3, 5 and 7 cannot be factored into smaller positive whole numbers. Their distribution appears irregular, but the Riemann Hypothesis connects their pattern to the zeros of the Riemann zeta function. The Clay Mathematics Institute states the conjecture as the claim that all its non-obvious zeros have real part one-half. A proof would sharpen many results in number theory that currently depend on the hypothesis being true.
Computations have checked the hypothesis for the first 10 trillion solutions, according to the Clay Institute, but checking a finite number of cases cannot prove a statement about infinitely many cases. Alpöge’s claim is therefore not established by the number of examples examined; it depends on whether the proposed argument is valid and proves the precise conjecture.
Formalization in Lean could help make an argument mechanically checkable, but a successful software check is not the same as broad mathematical acceptance. The code must be shown to encode the intended theorem correctly, and mathematicians must understand or otherwise rigorously validate the reasoning. The source material says Terence Tao described this submission as an “(enormous) Lean-only proof,” while Andrew Granville called the released note “garbled.” These comments address the submission and its presentation; they do not establish whether the proof is correct.
“I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis.”
— Levent Alpöge, in his November 20 note
Key Proof Checks Remain Open
The central unresolved issue is whether the Lean archive proves the exact statements Alpöge claims. The source material reports that an independent compilation succeeded, while the comparator run was still awaited. It does not provide the outcome of that test or a completed review by mathematicians who have verified the argument from its foundations.
It is also unclear which Claude model was involved, how much of the argument was generated by the system, and how Alpöge’s mathematical guidance shaped the result. The note says Claude contributed heavily, but those details are the author’s account. Anthropic had not announced the claim on its news page as of November 21, and Nature reported that the company did not immediately respond to its request for comment.
No journal has published the proof, and no mathematical body has accepted it. The Clay Institute’s prize process requires publication in a refereed journal, at least two years of scrutiny, and general acceptance among mathematicians. Accordingly, the prize clock has not started, and the claim should not be described as a solved problem.
Independent Review and Formal Checks
The next immediate milestone is the reported comparator run, followed by scrutiny of whether the Lean archive formalizes the intended conjectures without relying on hidden assumptions or proving a different statement. Independent mathematicians will need access to the code and time to examine its structure, dependencies and mathematical meaning.
If reviewers find the argument sound, the work would still need to move through publication and broader mathematical review. Under the Clay Institute’s stated rules, a qualifying proof must appear in a refereed journal and withstand at least two years of scrutiny before general acceptance can be recognized for the prize. If checks identify errors or gaps, Alpöge or other researchers may need to revise the formalization or withdraw the claim. No review timetable or journal submission is specified in the available material.
Key Questions
Has the Riemann Hypothesis been proved?
Not on the evidence currently reported. Alpöge has claimed a proof, but it has not been accepted by mathematicians, and key verification remains incomplete.
What did Claude reportedly do?
Alpöge says he guided the strategy and Claude carried out much of the technical work. That division of work is Alpöge’s account; the note does not identify the Claude model used.
Does the Lean code prove the claim because it compiled?
No. Compilation indicates that Lean accepted the code under its formal rules, but reviewers still need to check that it represents the intended theorem and that the argument establishes the claimed result. A comparator test was reportedly still pending.
What is the Riemann Hypothesis about?
It concerns the locations of the non-obvious zeros of the Riemann zeta function, asserting that they all have real part one-half. The conjecture has important consequences for understanding the distribution of prime numbers.
Has the $1 million prize been awarded?
No. The Clay Mathematics Institute’s process requires journal publication, at least two years of scrutiny and general acceptance. Those conditions have not been met, so the prize process has not begun.
Source: ThorstenMeyerAI.com
