Anthropic mathematician claims AI-assisted proof of Riemann hypothesis

Covered by 19 sources
Read onBBC· 3 min
Levent Alpöge of Anthropic asserts he and the AI model Claude have proven the Riemann hypothesis, formalizing the work in Lean code. The claim awaits independent verification and has drawn scrutiny for its brevity and lack of traditional exposition.

What happened

On November 20, 2026, Levent Alpöge, a mathematician at Anthropic, announced that he and the AI model Claude had produced a formal proof of the Riemann hypothesis, a longstanding unsolved problem in mathematics. The proof was formalized in Lean code and published on Alpöge's personal website. This announcement came shortly after Anthropic's IPO. The proof has not yet been independently verified or accepted by the mathematical community.

Key details

  • The Riemann hypothesis is a Millennium Prize Problem with a $1 million reward for a correct solution.
  • Alpöge credits Claude, Anthropic's AI model, for key insights in the proof.
  • The proof was developed over six weeks and formalized in the Lean proof assistant system.
  • The Clay Mathematics Institute has not yet accepted the proof.
  • A separate, shorter proof was released by Stanford's Alessandro Masullo after Alpöge's announcement.

Why it matters

Mathematicians, AI researchers, and institutions like the Clay Mathematics Institute now face the task of verifying the validity of Alpöge and Claude's proof. If accepted, this would resolve a 167-year-old problem and demonstrate a new capability for AI in mathematical research. The outcome could affect the credibility and future role of both human and AI mathematicians in solving major scientific challenges.

Also covered by 18 sources

More on this story

  1. Hours after Anthropic claims a Claude proof of the Riemann hypothesis, Stanford releases an all-human competing draft
    · The Decoder

More in AI Research