Anthropic mathematician Levent Alpöge released a Lean-compilable proof of the Riemann hypothesis Friday, the same week his employer became the first publicly traded frontier AI lab.
Alpöge posted the formal verification files and a four-page explanation to GitHub and Zulip, describing a “stream-of-consciousness jaunt” through the final argument with Anthropic’s Claude model, TechCrunch reported. The archive spans roughly 239 megabytes and 175,187 source files, Interesting Engineering reported. Independent researchers, including MIT mathematician Andrew Sutherland, have confirmed the code compiles against Mathlib, Lean’s standard mathematics library, according to TechCrunch.
The Riemann hypothesis, posed by Bernhard Riemann in 1859, conjectures that every nontrivial zero of the zeta function lies on the critical line where the real component equals one-half. A proof would resolve a Millennium Prize problem carrying a $1 million award from the Clay Mathematics Institute, though the Institute requires publication in an approved peer-reviewed journal and two years of broad acceptance before consideration, Interesting Engineering noted.
Alpöge’s claim goes further: he asserts a proof of the Generalized Riemann Hypothesis, which extends the conjecture to Dirichlet L-functions.
Chris Birbeck, a number theorist at the University of East Anglia, wrote on Zulip that the cohomology bound relies on an automorphy-lifting argument using Galois representations attached to regular algebraic, polarized cuspidal automorphic representations of GL_n over CM fields, along with base change, Jacquet–Shalika, Clozel’s purity, Bernstein–Zelevinsky, the main theorem of complex multiplication, and Deligne–Ribet, TechCrunch reported.
Compilation alone does not establish that the intended conjecture has been proved. Researchers must still verify that the formal definitions match the classical statement and that no unsupported assumptions remain, Interesting Engineering reported.
Alpöge acknowledged the gap, writing that he felt “a bit guilty about releasing the proof without a truly human-intelligible explanation” but “could not live” if he waited, according to TechCrunch. He has committed to producing a narrative proof.
Simultaneously, Stanford professor Ravi Vakil, president of the American Mathematical Society, posted a draft proof of the Generalized Riemann Hypothesis by Stanford postdoc Alessandro Masullo, noting it is “all human, with no AI,” TechCrunch reported. Masullo described the 12-year effort as a message that mathematicians “are still here. We are very much alive. We are not going anywhere,” according to TechCrunch.
The Clay Mathematics Institute has not commented on either submission.
Related reading