An artificial intelligence researcher at Anthropic claimed on Friday to have solved the Riemann hypothesis using the company’s Claude chatbot.
“I believe I have, working with Claude, discovered a proof of the Riemann hypothesis,” Levent Alpoge wrote in a four-page note released Friday.
A Century-Old Math Mystery
Widely considered the premier open problem in pure mathematics, the hypothesis focuses on prime number distribution and was initially put forward in 1859 by the German mathematician Bernhard Riemann.
In 2000, the Clay Mathematics Institute designated it as one of seven Millennium Prize Problems, providing a $1 million reward for anyone who solves one.
Per the rules set by the institute, any valid solution needs to appear in a peer-reviewed journal and endure two years of rigorous evaluation.
Employed as a mathematician at Anthropic, Levent Alpöge earned his PhD at Princeton University in 2020 under the guidance of Fields Medalist Manjul Bhargava, and subsequently served as a junior fellow in the Harvard Society of Fellows from 2021 to 2025.

Having remained unsolved for 167 years, the Riemann Hypothesis stands as number theory’s classic challenge and one of mathematics’ most celebrated open questions. Introduced in 1859 by Bernhard Riemann—who also gave the problem its title—the theory posits that prime numbers are governed by underlying rules rather than being random, maintaining that all significant equation solutions rest on a specific vertical straight line.
Releasing Code Without a Paper
This really is a historic moment, Levent Alpöge wrote after he had compiled the result in Lean.
Holy cow. Anthropic is claiming to have solved the Riemann Hypothesis!!! If true, huge.#RiemannHypothesis https://t.co/80OxaFhmP6https://t.co/X7XMnFp7OK
“The world has changed… I know the mathematical community will have all sorts of reactions, but for me this is a triumph for humanity, a day worth celebrating,” Andrew Sutherland, MIT mathematician, added.
Months or even years may pass before the formal code completes the traditional line-by-line academic peer review process. This leaves the proof’s correctness unverified until that time.
Broader Mathematical Implications
According to Alpoge, the technique also settles the generalized Riemann hypothesis—an expanded variation of the problem. This validates a massive extension of Riemann’s original conjecture that supports the entire mathematical universe through vast layers of hidden structure.
If this proves to be true, it will be a monumental breakthrough. An ode to mathematics, to AI, and above all, to human ingenuity. After all, there is no such thing as AI born out of AI only.https://t.co/3NJmjnIJcD
“If this proof is correct and understood, it would complete a dictionary of correspondences between different branches of mathematics that until now was only partial. It would be like a Rosetta Stone,” Hector Pasten, a mathematician at the Pontifical Catholic University of Chile, told Scientific American.
Concordia University professor and Google DeepMind visiting Carlo Pagano said that the proof introduces fresh concepts “very likely to open up entirely new ways of thinking in number theory.”
For 167 years, the Riemann hypothesis eluded mathematicians. One day after the company’s IPO, an Anthropic employee says Claude helped discover a proof. https://t.co/qvhaizPoMh via @WSJ
A Crowded Day for Proofs
Hours after the AI-assisted claim appeared, Stanford mathematician Ravi Vakil, who is president of the American Mathematical Society, published a blog post. He announced that another mathematician, Alessandro Masullo, has a proposed proof of the generalized Riemann hypothesis without any AI.
Alpöge said he had consulted Anthropic chief executive Dario Amodei before announcing the result. He acknowledged that its timing could be viewed cynically, coming in the same week that Anthropic was floated on the Nasdaq in its initial public offering.

Anthropic revealed in an August internal study that an unreleased research model of Claude stopped short of proving the Riemann Hypothesis, yet successfully increased the known lower bound for non-trivial zeros situated on the critical line from 41.6% to 67.2%.
Alpöge, who earned his doctorate at Princeton under Fields Medalist Manjul Bhargava and served as a junior fellow at Harvard from 2021 to 2025, said in July that he had also solved the Jacobian conjecture using Claude.
He noted that he had been working on the Navier-Stokes problem with NYU professor Tristan Buckmaster, a route he said OpenAI later pursued with Codex and Claude to produce a widely discussed but unverified solution. Alpöge added that he texted Anthropic chief executive Dario Amodei about the Riemann hypothesis result late Sunday night and held off publication until after the company’s initial public offering on November 19, 2026.
Alpöge said that his work with Claude took 41 days, spanning from October 6 to November 16.