Anthropic Mathematician Claims Proof of Riemann Hypothesis
An Anthropic employee has released a preprint claiming a general proof of the Riemann Hypothesis, a 167-year-old unsolved problem. The claim has sparked intense debate on Hacker News and blogs by Terry Tao and Scott Aaronson, with a competing human solution also emerging.
On November 20, 2026, Anthropic mathematician Levent Alpöge released a preprint claiming a proof of the Generalized Riemann Hypothesis (GRH) using Anthropic's Claude LLM, with the proof coded in the Lean formal verification language. The announcement, coming a day after Anthropic's IPO and amid a year of AI math breakthroughs, has sparked intense debate on Hacker News and blogs by Terence Tao and Scott Aaronson, while a competing human solution by Stanford postdoc Alessandro Masullo also emerged.
-
Proof details and verification status
Alpöge's proof consists of 17.4 million lines of Lean code, which MIT mathematician Andrew Sutherland independently compiled successfully. The proof targets the Generalized Riemann Hypothesis, a broader version of the original 1859 conjecture. Lean verification, while not infallible due to past kernel bugs, has so far not required retraction of any AI-certified result, according to Scott Aaronson. [1] [3] [4]
-
Competing human solution
Stanford professor Ravi Vakil posted a human-written proof of GRH by postdoc Alessandro Masullo, the result of 12 years of work. Masullo's proof is not formalized in Lean, making verification uncertain. In his introduction, Masullo wrote, 'if any message at all this work is meant to send to the mathematician who feels discouraged by the pace and tone of the current climate, it is this: we are still here.' [5]
-
Mathematical and institutional context
The Riemann Hypothesis, first proposed in 1859, is one of the Clay Mathematics Institute's Millennium Prize problems with a $1-million bounty. Progress had been rare until October 2026, when OpenAI's math results made significant headway on prime structure. Alpöge's proof, if correct, would complete a 'Rosetta Stone' between branches of math, according to mathematician Hector Pasten. [1] [4]
-
Reactions from leading mathematicians
Terence Tao noted on his blog that the GRH proof is the least 'digested' of three announced results, with no human-readable path yet. Scott Aaronson bet the proof will stand, citing no retracted Lean-certified AI results. Andrew Sutherland called it 'a human triumph' and 'a day for celebration,' while acknowledging the need for time to process the implications. [1] [3] [4]
-
Timing and related announcements
Alpöge's announcement came the day after Anthropic's record-breaking IPO. He also released a separate proof related to the Navier-Stokes equations with NYU mathematician Tristan Buckmaster. Alpöge described messaging Anthropic CEO Dario Amodei the previous Sunday night to ask how to announce the finding, calling the timing 'horrible.' [1] [5]