An Anthropic mathematician has announced a claimed proof of the Riemann hypothesis with the help of the AI assistant Claude. If independently verified, this breakthrough would solve one of mathematics’s most challenging historical puzzles.
The Claim and the Millennium Prize
Levent Alpöge published a document on his website detailing the solution, which he states was developed alongside Claude. In his statement, Alpöge credits Claude as integral to the process and notes that the method also extends to the Generalized Riemann Hypothesis and the Extended Riemann Hypothesis. He texted Anthropic CEO Dario Amodei about the discovery late Sunday night, holding off publication until after the company’s IPO on Thursday.
The Riemann hypothesis is one of seven Millennium Prize problems, which carry a $1 million bounty for each solved challenge through the Clay Mathematics Institute. The problem, first proposed in 1859, focuses on prime number distribution and function zeroes. Under institute rules, a proposed proof must be published and undergo at least two years of scrutiny to receive general acceptance by the global mathematics community.
Verification and AI in Mathematics
Alpöge’s solution has not yet been independently verified or publicly accepted by the institute, and Anthropic has not commented on the announcement. In August, Anthropic stated that it was working on the Riemann hypothesis with Claude, though it expressed skepticism at the time that the techniques would yield a proof.
This development follows other recent AI milestones in mathematics, including OpenAI’s September claim to have solved the Navier-Stokes problem. That OpenAI effort involved millions of dollars and thousands of agents, though it faced controversy and allegations regarding the use of unpublished research from outside mathematicians. Alpöge previously stated in July that he had also solved the Jacobean conjecture using Claude, as the rapid influx of AI-driven mathematical findings reshapes the field.








