Overview
- Levent Alpöge, an Anthropic researcher, posted a four‑page statement on Friday saying he and the company’s Claude model produced a proof of the Riemann Hypothesis and related generalizations.
- Alpöge released the work as a massive Lean formalisation rather than a conventional paper, publishing hundreds of thousands of pages and more than 100,000 source files that encode each logical step for machine checking.
- MIT mathematician Andrew Sutherland and other independent reviewers have compiled Alpöge’s Lean archive without errors and reported that several principal theorems depend only on Lean’s standard axioms, though additional verifications are still running.
- Alpöge explicitly credits Claude with central contributions, says he consulted Anthropic CEO Dario Amodei before announcing, and published the claim immediately after Anthropic’s IPO, a timing some commentators have questioned.
- The Clay Mathematics Institute has not accepted the result and its rules require formal publication and at least two years of community scrutiny before considering the Millennium Prize, so full human review and explanatory writeups remain the next essential steps.