AIToday
Large Language ModelsAI Safety & AlignmentAI WeeklyPublished: Nov 21, 2026, 19:01 JST

Anthropic's Levent Alpöge claims Riemann Hypothesis proof with Claude

Send on LINE
Anthropic's Levent Alpöge claims Riemann Hypothesis proof with Claude

3 Key Points

  1. What happened

    Levent Alpöge, a mathematician employed at Anthropic, said he had proven the Riemann Hypothesis working with Claude, posting it as Lean code that runs to 175,187 source files and 239MB, with no human-readable paper.

  2. Why it matters

    No outside mathematician has confirmed the work and Anthropic has issued no statement; number theorist Andrew Granville called Alpöge's four-page sketch 'garbled,' so the claim could still fall at formal review.

WHO IT HITSResearch mathematicians who evaluate proofs and prize claims face the task of checking a Lean-coded argument rather than a conventional paper, while Anthropic and the Clay Mathematics Institute must decide how to treat a claim that carries a million-dollar bounty.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, every morning.

Context & Analysis

The claim rests on a 167-year-old problem that Bernhard Riemann first proposed in 1859, and it is one of the few carrying a million-dollar bounty from the Clay Mathematics Institute. Alpöge framed the work in unusually personal terms, writing that it is natural to feel this is an end and, in his case, a loss of a principal guiding light. He also acknowledged releasing a complete argument before having mastered a massive neighbourhood around its mathematics, a division of labor between human and AI that sits at the center of how the claim is being received.

Hector Pasten of the Pontifical Catholic University of Chile told Scientific American that if the argument holds it completes a dictionary between different branches of math where the connections were only partial, something like a Rosetta Stone. That potential is what makes the verification gap consequential: Lean, the programming language that lets a machine verify each step of a mathematical argument, should check quickly, but it confirms logical validity rather than that the problem was correctly encoded, so human mathematicians still have to verify the definitions.

Andrew Granville's 'garbled' description of the four-page sketch is the first named expert reaction on record before formal review has begun, and Anthropic has issued no statement. The claim also emerged one day after Anthropic's $93B Nasdaq IPO, according to one outlet's account, which may shape how much attention the episode draws beyond the mathematics community.

FAQ
Does the proof also cover a broader problem than the Riemann Hypothesis?
Yes. Alpöge says the proof also settles the Generalized Riemann Hypothesis, expanding the claim beyond the standard Millennium Prize formulation.
How large is the proof file, and can a human read it?
It is posted as Lean code across 175,187 source files and 239MB of data, and no human-readable paper has been published alongside it.
What did the first named expert say about the proof?
Number theorist Andrew Granville called Alpöge's four-page sketch 'garbled' and said he should have prepared a clearer write-up before the announcement.
Send on LINE

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleNakaTaku launches NPC for physical AI components