Sunday, 22 November 2026, story 7 of 7
Lean communities certify Alpöge's Claude-assisted Riemann proof
A separate all-human Stanford draft has drawn reported problems; its announcer, Ravi Vakil, makes no claim it is correct.

What happened
The Lean and Mathlib communities say Levent Alpöge's Claude-assisted proof of the Riemann Hypothesis, reported here yesterday, is correct. Their joint statement says dozens of people reproduced the compilation, two tools confirmed the code proves Mathlib's own statement of the hypothesis, and several independent proof checkers accepted the exported proof.
Separately, Ravi Vakil, Stanford professor and American Mathematical Society president, posted an all-human draft proof of the Generalized Riemann Hypothesis by Alessandro Masullo. About eight hours later he added that people had reported problems, found with help from AI models, and said he makes no claim the strategy is correct.
What it means
The Stanford draft does not decide whether Alpöge's proof stands. The two are independent, and the Lean checks the communities describe run on the formal proof itself. If their account holds, the risk Scientific American raised, that the problem was coded wrongly, has been addressed by comparing the theorem against Mathlib's vetted statement.
What remains open is understanding. The statement says it looks forward to work that explains the new ideas, and Alpöge says he will write a human-readable version over the coming months. The statement names the Riemann Hypothesis only. Alpöge's claim to the Generalized version is not addressed in it. Watch whether that claim gets the same scrutiny, and whether the problems reported in Masullo's draft are resolved.
What to do: If you rely on AI-generated proofs, note how this one was trusted: independent checkers and a comparison against a vetted statement, not a reader's judgment.
Sources
- Hours after Anthropic claims a Claude proof of the Riemann hypothesis, Stanford releases an all-human competing draft, The Decoder
- Lean verification of the Riemann Hypothesis, leanprover-community.github.io
Earlier on this: Anthropic mathematician claims Claude-assisted Riemann hypothesis proof (21 November 2026).
Filed under Research. Published .
"What it means" is analysis, not reporting. Found an error? Write to daily@vibitely.com.