Skip to content
Theme
OnDevice Tools

AIChatbots & Assistants

Riemann Hypothesis proof claimed by Anthropic employee using Claude

An Anthropic mathematician claims he used the Claude AI to produce a proof of the Riemann Hypothesis, a famous unsolved problem dating back to 1859.

Illustration of a laptop screen showing a mathematical proof written in the Lean programming language

Key takeaways

  • Levent Alpöge of Anthropic says he used Claude to prove the Riemann Hypothesis.
  • The proof is written in Lean and is said to be over 250,000 pages long.
  • An Oxford mathematician told The New York Times he is 98 or 99 percent certain it is correct.
  • The Clay Mathematics Institute offers one million dollars for a verified solution.
In this article
  1. What happened
  2. The proof and the claim
  3. Expert reaction
  4. Why it matters

What happened

Illustration: Claude logo, Anthropic logo, AI, news and the labels Anthropic, Millennium Prize, Riemann Hypothesis

Anthropic employee Levent Alpöge says he used the Claude artificial intelligence system to produce a proof of the Riemann Hypothesis. Engadget reported the claim on 21 November 2026, citing the BBC. If verified, it would solve one of the most famous unsolved problems in mathematics.

Alpöge, a mathematician working for Anthropic, said he worked with Claude for six weeks to create the detailed proof. He described the problem as important to him since childhood. The proof was written in Lean, a programming language used to write and confirm mathematical proofs. Alpöge said the complete proof runs to more than 250,000 pages.

The proof and the claim

The Riemann Hypothesis was first proposed by Bernhard Riemann in 1859. In simple terms, it suggests that prime numbers appear along the number line in a meaningful order. Prime numbers are whole numbers that can only be divided by themselves and one to produce another whole number. Scientific American describes the hypothesis as the most important unsolved problem in math. The Clay Mathematics Institute lists it as one of its Millennium Prize Problems and offers one million dollars for a correct solution.

Expert reaction

James Maynard, a mathematician at the University of Oxford in England, told The New York Times that he believes the discovery is accurate. He estimated his confidence at 98 or 99 percent. Other mathematicians have not yet verified the proof, and it could take time to check a document of this length. OpenAI also shared progress on the Riemann Hypothesis in October, according to the report.

Why it matters

If the proof holds up, it would mean an AI system solved a problem that has resisted mathematicians for 167 years. The Clay Mathematics Institute prize of one million dollars would go to the first correct solution. The claim adds to recent AI advances in mathematics, though independent review is still needed before the result is accepted.

Sources

  1. Anthropic employee claims to have proven the most famous math problem using ClaudeEngadget

Tags

Spotted a mistake? See our corrections policy. Read how News by OnDevice works in our editorial policy.