An Anthropic mathematician says he has proved the Riemann Hypothesis, one of the most famous unsolved problems in mathematics, with help from the company's Claude AI model. The claim has not been independently verified. Scientific American, OfficeChai and SSBCrack all reported the claim on Friday.
Levent Alpöge, a mathematician who works at Anthropic, is the person named in the reports. The note opens with the line: “I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis.” Scientific American said Alpöge wrote this in a statement on his website.
OfficeChai described the situation differently. It reported that a four-page document titled “Remark” appeared on Alpöge's personal website with no author's name, and that he had not shared it on social media or elsewhere. The sources therefore differ on how the note was presented, and all three outlets say the claim still needs checking.
Bernhard Riemann proposed it in 1859. According to OfficeChai, it concerns the zeros of the Riemann zeta function, which encodes how prime numbers are spread among the integers. The hypothesis says all of the function's “nontrivial” zeros lie on a single vertical line in the complex plane. Scientific American said it suggests that the seemingly random placement of primes follows a hidden order.
It is one of seven Millennium Prize Problems, with a $1 million reward from the Clay Mathematics Institute, OfficeChai reported. Thousands of results in number theory are currently proved only on the assumption that the hypothesis is true.
The note says it also proves the Generalized Riemann Hypothesis, which covers a broader family of related functions, according to OfficeChai and SSBCrack. No paper has been published. The argument was released only as code in Lean, a programming language in which a computer can check every logical step of a proof. OfficeChai said the Lean files sit on the same website in 14 compressed archive parts.
The author wrote that the argument is too long and unfinished to write up in readable form, and that a human-readable version is promised later. The note also includes a 20-step timeline that it says Claude compiled, with the last step, the Generalized Riemann Hypothesis, dated Nov. 16, OfficeChai reported.
OfficeChai said the strategy borrows from “function fields,” a setting in which the equivalent of the hypothesis has already been proved. One route there, Stepanov's method, depends on derivatives, which have no natural counterpart for ordinary integers. The note's central idea is a workaround that uses a long chain of “theta lifts” instead, and it credits Claude with the key step. The author describes the work as a joint effort and says Claude far outpaced him near the end.
The note is also frank about its limits, OfficeChai said. It calls the timeline rough, warns that the argument may contain odd or suboptimal choices and says some steps hold open questions or the author's intuitions.
Not yet. OfficeChai reported that nobody outside the author's circle has confirmed that the Lean archive builds or that its final theorem says what the Riemann Hypothesis says. SSBCrack added that posting a Lean file does not by itself prove the conjecture: specialists must confirm that the formal argument is sound, that it proves the intended statement and that no unsupported assumptions remain. Scientific American said experts should be able to verify the coding of the hypothesis fairly quickly.
If the proof holds up, it would settle a problem that has resisted solution for more than a century and a half, SSBCrack noted. Scientific American said the result would have no immediate real-world applications and would not change global banking or how the internet runs.
The claim also lands amid growing debate over what AI can do in research mathematics. SSBCrack recalled that Alpöge earlier shared a Claude-generated solution to the Jacobian conjecture, and that a collection of results produced with an OpenAI system was reported to include progress on questions about prime numbers. Alpöge wrote that the prospect could bring mixed feelings for mathematicians, but that he sees it as the dawn of something larger across science.
Around Prague, based on reporting by Scientific American, OfficeChai, SSBCrack