EDITION OF SAT 21 NOV 2026
13 · TECHNOLOGY

Independent checks support Claude-assisted proof of Riemann hypothesis

publishers 42articles 53first reported 20 Nov, 00:00 UTCdeveloping since 20 Nov · 3 editionstop 25 #13

Independent mathematicians have successfully compiled and checked a Claude-assisted proof of the Riemann hypothesis released by Anthropic mathematician Levent Alpöge, according to Scientific American. The Lean and Mathlib communities said in a joint statement that both the theorem’s formulation and its proof were correct, while MIT mathematician Andrew Sutherland said the Lean kernel had independently rechecked it using only three standard axioms.

Alpöge announced on November 20 that he had developed the proof while working with Anthropic’s Claude AI model. He also said the work covered the generalized Riemann hypothesis. Rather than publishing a conventional paper, he released the argument as about 17 million lines of Lean code, allowing its logic to be checked step by step by the proof assistant.

The independent checks addressed whether the code compiled, whether it stated the actual Riemann hypothesis and whether it used any hidden shortcuts or altered assumptions. Researchers had initially cautioned that verification and human understanding of the argument could take time. Mathematicians are now translating and examining the code to determine how the proof works and which ideas are new.

First proposed by Bernhard Riemann in 1859, the hypothesis concerns the distribution of prime numbers and is one of the Clay Mathematics Institute’s seven Millennium Prize Problems. The institute offers $1 million for an accepted solution, subject to its publication, waiting-period and general-acceptance requirements.

On the same day, Stanford mathematician Alessandro Maria Masullo also released a separate manuscript described by American Mathematical Society president Ravi Vakil as an entirely human effort.

HOW THIS STORY WAS MADE

Written from 8 articles, 6 with the publisher's own text; 8 independent newsrooms once syndicated copies count as one; this version written 24 h after the record first saw the story; the editor revised it.

Publishers
42
Source articles
53
Given to the writer
8, 6 with the article's own text
Independent newsrooms
8
This version written
24 h after the record first saw the story
Second model (editor)
revised
Publication gate
passed
Human review
none

WHAT THE EDITOR DID

The second model checked the writer's draft against the same sources and changed it before publication. Struck words were in the draft; underlined words are its corrections.

WHAT IT FLAGGED editor revise: Delete the unsupported sentence about the separate manuscript’s verification status. - "Its verification status was not reported.": This is an unsupported statement about the absence of reporting rather than a fact established by the sources.

THE DRAFT AND THE CHANGES

Independent checks support Claude-assisted proof of Riemann hypothesis

Independent mathematicians have successfully compiled and checked a Claude-assisted proof of the Riemann hypothesis released by Anthropic mathematician Levent Alpöge, according to Scientific American. The Lean and Mathlib communities said in a joint statement that both the theorem’s formulation and its proof were correct, while MIT mathematician Andrew Sutherland said the Lean kernel had independently rechecked it using only three standard axioms. Alpöge announced on November 20 that he had developed the proof while working with Anthropic’s Claude AI model. He also said the work covered the generalized Riemann hypothesis. Rather than publishing a conventional paper, he released the argument as about 17 million lines of Lean code, allowing its logic to be checked step by step by the proof assistant. The independent checks addressed whether the code compiled, whether it stated the actual Riemann hypothesis and whether it used any hidden shortcuts or altered assumptions. Researchers had initially cautioned that verification and human understanding of the argument could take time. Mathematicians are now translating and examining the code to determine how the proof works and which ideas are new. First proposed by Bernhard Riemann in 1859, the hypothesis concerns the distribution of prime numbers and is one of the Clay Mathematics Institute’s seven Millennium Prize Problems. The institute offers $1 million for an accepted solution, subject to its publication, waiting-period and general-acceptance requirements. On the same day, Stanford mathematician Alessandro Maria Masullo also released a separate manuscript described by American Mathematical Society president Ravi Vakil as an entirely human effort. Its verification status was not reported.

Written by a language model from the sources above, then checked by a second model that may only cut, attribute or correct. How it works →

NAMED IN THE COVERAGE

OUTLETS CARRYING THE STORY, RUN BY RUN

0 22 45 21 Nov 22 Nov 21 Nov 00:16 UTC · Anthropic Mathematician Claims Claude-Assisted Proof of Riemann Hypothesis 21 Nov 06:14 UTC · Competing AI-assisted and human proofs claim to solve Riemann hypothesis 21 Nov 12:13 UTC · Anthropic mathematician claims Claude-assisted proof of Riemann hypothesis 21 Nov 18:13 UTC · Independent checks support Claude-assisted proof of Riemann hypothesis

WHAT HELD, WHAT CAME LATER

From the first reports to the latest

  • Riemann Hypothesis · 8 outlets
  • Bernhard Riemann · 7 outlets
  • Andrew Sutherland · 2 outlets

Entered the story later

  • Ravi Vakil · from 20 Nov 19:32, 2 outlets
  • Alessandro Masullo · from 20 Nov 19:32, 2 outlets
  • Kevin Buzzard · from 21 Nov 12:49, 2 outlets

HOW THIS STORY DEVELOPED: EVERY HEADLINE, EVERY OUTLET →