Anthropic Researcher Claims AI-Assisted Proof for the Riemann Hypothesis
A researcher at Anthropic has published a note claiming to have used the company’s AI software, Claude, to develop a proof for the century-old Riemann hypothesis.
- Anthropic researcher Levent Alpöge claims to have used Claude AI to prove the Riemann hypothesis.
- The Riemann hypothesis is a 150-year-old math problem concerning prime numbers and a Millennium Prize Problem.
- MIT scientist Andrew Sutherland and other mathematicians have reportedly verified the work on their own systems.

A mathematician and researcher at Anthropic PBC has stated that he used the company's artificial intelligence software, Claude, to formulate a proof for the Riemann hypothesis. The long-standing number-theory problem, which concerns the distribution of prime numbers, has remained unsolved for more than 150 years and is recognized as one of the Millennium Prize Problems. The researcher, Levent Alpöge, published a note detailing his findings on his personal website and linked to supporting code written in the Lean programming language.
Verification and Academic Response
Before releasing the findings, Alpöge contacted Anthropic Chief Executive Officer Dario Amodei on Sunday night to discuss the public announcement. Anthropic declined to comment on the development, and Alpöge did not respond to requests for comment. If independently confirmed, this would represent the second Millennium Prize Problem potentially resolved using artificial intelligence, following an announcement by OpenAI regarding the Navier-Stokes problem.

Independent validation of the work has begun within the academic community. Andrew Sutherland, a senior research scientist at the Massachusetts Institute of Technology, stated that he verified the proof on his own computer and noted that at least three other professional mathematicians had completed verifications on multiple systems. Noah Giansiracusa, an associate professor of mathematics at Bentley University and a faculty associate at Harvard University, stated that there is a strong likelihood the solution is valid and airtight.
- Levent Alpöge published his proof and Lean programming code online.
- MIT senior research scientist Andrew Sutherland verified the work independently.
- Anthropic declined to comment on the announcement.
Spotted an error? · Our editorial policy
UP NEXTSarfaraz Khan Details Emotional Toll of Dressing-Room Leak Allegations and Gambhir’s SupportOther angles · #Anthropic
Timeline →- NewsAnthropic Shares Pull Back Toward IPO Price Following Record Nasdaq Debut
- NewsAnthropic Stock Falls 10% Following Record-Breaking Nasdaq Debut
- NewsAnthropic Secures Record-Breaking $81.3 Billion IPO and $1.8 Trillion Valuation
- NewsAnthropic Enters Public Market with Record-Breaking Valuation Amid Internal Safety Debates
More from Technology

Grand Theft Auto VI Becomes Largest Entertainment Launch in History

Trump Backs Jensen Huang’s Return to Nvidia Amid Board Resistance and Intelligence Scrutiny

Donald Trump Publicly Supports Reinstatement of Former Nvidia CEO Jensen Huang
