How An Anthropic Mathematician Used Claude On The Riemann Hypothesis
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: How An Anthropic Mathematician Used Claude On The Riemann Hypothesis on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the little things that make your day delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

Anthropic mathematician Levent Alpöge says he and Claude found a proof of the Riemann Hypothesis and its generalized form, releasing a short note and a large Lean code archive. Early reports say the archive compiled, but independent checks are incomplete, and mathematicians have not accepted the proof.

Levent Alpöge, a mathematician who works at Anthropic, said in a note posted November 20 that he believes he and the company’s Claude AI found a proof of the Riemann Hypothesis, one of mathematics’ best-known unsolved problems. He released a short explanation and a large archive of Lean code, but no mathematical body has accepted the result, and key checks are still pending.

Alpöge’s four-page note says the approach also extends to the Generalized Riemann Hypothesis, a broader family of conjectures. The note is not a full, readable proof: it sketches some ideas, gives a timeline, and links to a formalization divided into 14 parts. The formalization is written in Lean, software that checks whether mathematical arguments follow from stated definitions and rules.

Alpöge says he chose to publish the Lean archive rather than a readable version, which he described as too unwieldy. He says a human-readable document he produced would exceed 250,000 pages. Computer scientist Scott Aaronson put the archive at about 17.4 million lines of Lean code. These figures describe the submitted material, not an independent assessment of its correctness.

The division of work is known only through Alpöge’s account. He says he directed the strategy and that Claude handled much of the technical work, writing that “Claude far outpaced me near the finish line.” He also says he had “little mastery” of one of the main areas used in the proof. The note does not identify which Claude model was involved. Anthropic had not announced the claim on its news page as of November 21; Nature reported that the company did not immediately respond to its request for comment.

At a glance
reportWhen: Claim posted November 20, 2026; review…
The developmentLevent Alpöge published a note saying he believes he and Claude proved the Riemann Hypothesis, along with a Lean archive that mathematicians are beginning to examine.

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

0 1/2 1 real part

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

  1. 1Claim postedA 4-page note plus Lean code, November 20, 2026✓ done
  2. 2Code compiles elsewhereIndependent compile reported by an MIT mathematician✓ reported
  3. 3Comparator runRight statement? Only standard axioms? Several Lean kernels?… pendingWe are here · November 21, 2026
  4. 4Humans understand itA readable proof that experts can follow○ open
  5. 5Journal publicationA refereed mathematics journal of worldwide repute○ open
  6. 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

Why a Proof Would Matter

The Riemann Hypothesis concerns the locations of special zeros of the zeta function and is closely tied to how prime numbers are distributed. A proof could settle many results in number theory that are currently known only conditionally. The Clay Mathematics Institute lists the hypothesis among its Millennium Prize Problems, each associated with a $1 million prize.

The immediate significance, however, is not that the problem has been solved, but that a claim backed by a large machine-checkable formalization is being scrutinized. Lean can verify that a coded argument follows its formal rules; that does not by itself tell mathematicians whether the formalized result matches the intended theorem or whether the setup has a flaw. The distinction matters when assessing both the mathematics and the role AI played.

Products Worth Considering

OMRON Complete 2-in-1 Upper Arm Blood Pressure Monitor with EKG

OMRON Complete 2-in-1 Upper Arm Blood Pressure Monitor with EKG

Garmin Forerunner 265S Running Smartwatch, 42 mm, Whitestone/Neo Tropic

Garmin Forerunner 265S Running Smartwatch, 42 mm, Whitestone/Neo Tropic

As an affiliate, we earn on qualifying purchases.

From Conjecture to Lean Code

Bernhard Riemann introduced the hypothesis in 1859. In plain terms, it says that all the zeta function’s “non-obvious” zeros have real part one-half. The Clay Mathematics Institute says the hypothesis has been checked for the first 10 trillion solutions. Such checks test many individual cases, but cannot establish a claim about infinitely many cases.

Formal proof software offers a different kind of check: a proof must be translated into small, explicit steps that the system can verify. Alpöge’s release therefore puts the code, rather than a conventional mathematical paper, at the center of the initial review. The Clay prize rules also mean that a claim does not qualify simply because code compiles: the work must be published in a refereed journal, undergo at least two years of scrutiny, and gain general acceptance among mathematicians.

“I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis.”

— Levent Alpöge, in his November 20 note

Checks Still Underway

Early technical feedback is not a verdict on the theorem. 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; the test is intended to check whether the formalized result proves the intended statement rather than a different or inadvertently weakened one. The available source material does not report the outcome of that check.

Other questions remain open: whether specialists can follow and validate the mathematical strategy, whether the formalization contains an error, and how much of the work was generated by Claude versus directed or checked by Alpöge. Andrew Granville called the released note “garbled,” while praising or criticizing a short sketch does not itself settle the code’s validity. No journal publication or broad mathematical acceptance has been reported. A separate claim from a Stanford mathematician was also reported, but the supplied material gives no detail that would allow its merits to be compared.

Independent Review and Publication

The next step is independent examination of the Lean archive, including the pending comparator test and review by mathematicians who work in the relevant areas. If the code and the link between its formal statement and the Riemann Hypothesis withstand those checks, the work would still need a readable account and the usual scholarly review. Under the Clay Institute’s rules, the prize process requires publication in a refereed journal, at least two years of scrutiny, and general acceptance by mathematicians.

For now, the public record is a claim, a short note, and code under review. The source material does not establish a timeline for completing the checks, identify the Claude model used, or report an Anthropic response. Those points may become clearer as reviewers publish their findings.

Key Questions

Has the Riemann Hypothesis been proved?

Not according to the information available. Alpöge says he believes he and Claude found a proof, but mathematicians and a mathematical institution have not accepted it.

What did Alpöge release?

He posted a four-page note and a Lean formalization divided into 14 parts. The note sketches the claim; it does not provide a conventional, readable proof.

What does the Lean code check?

Lean checks whether a proof’s formal steps follow from its definitions and rules. Reviewers still need to establish that the formalized result is the intended theorem and that the overall argument is sound.

What role does Alpöge say Claude played?

Alpöge says he directed the strategy and Claude carried out much of the technical work. That account has not been independently verified, and the note does not identify the Claude model used.

Could this claim qualify for the Clay prize now?

No. The Clay Mathematics Institute’s rules require publication in a refereed journal, at least two years of scrutiny, and general acceptance among mathematicians before a prize can be awarded.

Source: ThorstenMeyerAI.com

You May Also Like

The Stanford AI Index 2026 Audit: Reading the Field’s Annual Report Card With a Critic’s Pen

The Stanford AI Index 2026, released three weeks ago, offers a comprehensive report on AI progress, but its methodology and interpretive claims warrant critical review.

Critical Insights Into AI Sovereignty Testing Through The 24% Rule

An in-depth analysis of France’s SecNumCloud sovereignty rule and its impact on AI and cloud providers, highlighting the significance of ownership caps.

Agency Tips: Rebuilding Email Automation Flows During Platform Moves

Agencies can now use new tools to automate the rebuilding of email automation flows during platform migrations, reducing manual effort and errors.

The Internet’s AI Security Record, Examined Through Recent Headlines

Recent disclosures involving PHP, weak certificate keys, cloud data and software packages highlight separate security failures and urgent patching needs.