· FourWeekMBA
Riemann Hypothesis Proof Certified After 167 Years
Photo: Antoine Dautry on Unsplash
Levent Alpoge, an Anthropic mathematician, discovered a proof of the Riemann Hypothesis while working with Claude, which was certified by Lean and Mathlib on November 21. The formal proof consists of 17 million lines of Lean code. The proof also resolves the Generalized Riemann Hypothesis.
Levent Alpoge, an Anthropic mathematician, stated "I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis" in a note released on November 20. The claim represents a watershed moment in mathematics: the Riemann Hypothesis, posed 167 years ago, now has a certified proof.
The Lean and Mathlib communities certified on November 21, 2026 that Alpoge's 17-million-line Lean library correctly states and proves the Riemann Hypothesis, following a day of compilation, statement-matching, and proof-kernel checks. The Lean FRO and the Mathlib Initiative said in a joint statement: "We hereby certify that both the statement and the proof of the Riemann Hypothesis are correct".
Dozens of people compiled Alpoge's code and reproduced the same proof export, and two independent tools, called comparator and con-leche, confirmed that the theorem the library proves is exactly the Riemann hypothesis as Mathlib states it. The formal proof establishes that every nontrivial zero of the Riemann zeta function has real part ½, with independent checks by MIT mathematician Andrew Sutherland also covering the generalized hypothesis for Dirichlet L-functions.
Alpoge's proof ships as roughly 17 million lines of Lean code in 14 archives, with an informal write-up over 250,000 pages. The achievement marks a historic collaboration between AI and human mathematics, demonstrating how large language models like Claude can assist in solving century-old open problems when properly guided and rigorously verified.
Sources & credits
Original source: FourWeekMBA