Skip to content
ESCAPE MERIDIAN

Anthropic · Sep 4, 2026

Anthropic reports Claude produced a computer-checked proof of Fermat’s Last Theorem

Markdown · JSON · Original source

Anthropic’s article published September 4, 2026, reports that Claude produced a computer-checked proof of Fermat’s Last Theorem using Lean, a proof assistant that checks mathematical reasoning. Anthropic says the work followed a simplified version of Andrew Wiles’s proof and used Prove2Me, a platform for coordinating agents’ formalization work. Anthropic says the project took 11 days and produced 13 million lines of Lean code. Kevin Buzzard, whose team had worked on formalizing the theorem, described the result as an autoformalization achievement. Anthropic presents formalization—encoding proofs so software can check them—as a way to help reviewers assess mathematical work, while saying it should not replace an explanation written for people. The article argues that this approach could reduce the burden of checking future mathematical results. Whether it can do so broadly, and how much human review such proofs will still require, remain open questions.

Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.

AnthropicArticle · anthropic.comOpen original source

Original source

Anthropic · Article

Source details

Original source date: Sep 4, 2026 (2026-09-04T00:00:00.000Z).

Date basis: Date stated on the publisher’s page.

Added to Escape Meridian: 2026-10-29T17:09:15.864Z. First observed: 2026-10-18T11:31:21.756Z. Retrieved: 2026-10-28T04:47:51.643Z.

Referenced by

Recorded citations and public file attachments. A reference alone does not establish agreement or influence.

No public references recorded here yet.

Reference data (JSON)

Limits

Text as retrieved on October 28, 2026; originally published September 4, 2026. The page was modified after publication. The retrieved text may include later additions, which cannot be distinguished from the original article using the supplied evidence. Anthropic’s claims about the proof and its possible implications are reported by the publisher; this evidence does not independently assess them.

Open question

How much human review will formalized proofs of other major mathematical results still require?

Source-stated

Anthropic reports that Claude produced an end-to-end, computer-checked proof of Fermat’s Last Theorem in 11 days, working largely autonomously.

In 11 days, working largely autonomously, Claude produced the first end-to-end, computer-checked proof of FLT.

Source-stated

Anthropic says the proof contained 13 million lines of Lean code.

At 13 million lines of Lean code, Claude’s proof is over 5x the size of Mathlib

Source-stated

Anthropic says Prove2Me helped agents coordinate their work and reuse theorem descriptions during the formalization.

Maintaining a directed acyclic graph (DAG) of theorem statements that agents used to decide what proofs they should attempt next.

Source-stated

Anthropic argues that formalized proofs could help reduce the burden of reviewing new mathematical work, but says they should not replace explanations for human readers.

Although we do not think a formalized proof should replace a human-understandable exposition, it may be the only feasible way for the mathematical community to keep up with AI-generated contributions.

Source and contributor context

Anthropic · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Formal verification

Using software to check every step of a proof or a program against precise rules.

Anthropic

Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.

Claude

Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.

Lean

Lean is a programming language and proof assistant used to express mathematical statements and check formal proofs.

Editorial provenance

Version 1. Publication ID: 8bfd2c63-dd26-d2ab-c2f9-334719a262e7.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T17:09:09.675Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: d4fea46f337b87c238e5a0f0350f0a02ca2392422f778722079853bd95260ff1. Source text hash: cad4314b74c030e1bc0f85dc9da9c758da69ed3a1579bd494f1e6640af1c6150.

Revision and source-check history

Version 1 · 2026-10-29T17:09:15.864Z

Anthropic reports Claude produced a computer-checked proof of Fermat’s Last Theorem

Anthropic’s article published September 4, 2026, reports that Claude produced a computer-checked proof of Fermat’s Last Theorem using Lean, a proof assistant that checks mathematical reasoning. Anthropic says the work followed a simplified version of Andrew Wiles’s proof and used Prove2Me, a platform for coordinating agents’ formalization work. Anthropic says the project took 11 days and produced 13 million lines of Lean code. Kevin Buzzard, whose team had worked on formalizing the theorem, described the result as an autoformalization achievement. Anthropic presents formalization—encoding proofs so software can check them—as a way to help reviewers assess mathematical work, while saying it should not replace an explanation written for people. The article argues that this approach could reduce the burden of checking future mathematical results. Whether it can do so broadly, and how much human review such proofs will still require, remain open questions.

Original source

Anthropic · Article

Source details

Original source date: Sep 4, 2026 (2026-09-04T00:00:00.000Z).

Date basis: Date stated on the publisher’s page.

Added to Escape Meridian: 2026-10-29T17:09:15.864Z. First observed: 2026-10-18T11:31:21.756Z. Retrieved: 2026-10-28T04:47:51.643Z.

Limits

Text as retrieved on October 28, 2026; originally published September 4, 2026. The page was modified after publication. The retrieved text may include later additions, which cannot be distinguished from the original article using the supplied evidence. Anthropic’s claims about the proof and its possible implications are reported by the publisher; this evidence does not independently assess them.

Open question

How much human review will formalized proofs of other major mathematical results still require?

Source-stated

Anthropic reports that Claude produced an end-to-end, computer-checked proof of Fermat’s Last Theorem in 11 days, working largely autonomously.

In 11 days, working largely autonomously, Claude produced the first end-to-end, computer-checked proof of FLT.

Source-stated

Anthropic says the proof contained 13 million lines of Lean code.

At 13 million lines of Lean code, Claude’s proof is over 5x the size of Mathlib

Source-stated

Anthropic says Prove2Me helped agents coordinate their work and reuse theorem descriptions during the formalization.

Maintaining a directed acyclic graph (DAG) of theorem statements that agents used to decide what proofs they should attempt next.

Source-stated

Anthropic argues that formalized proofs could help reduce the burden of reviewing new mathematical work, but says they should not replace explanations for human readers.

Although we do not think a formalized proof should replace a human-understandable exposition, it may be the only feasible way for the mathematical community to keep up with AI-generated contributions.

Source and contributor context

Anthropic · Publisher

The publisher is identified here. No individual author has been inferred.

Source media

Media retrieval status: unchecked.

Terms and references

Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.

Formal verification

Using software to check every step of a proof or a program against precise rules.

Anthropic

Anthropic develops Claude and publishes research about AI capability, safety and how people use its systems.

Claude

Claude is Anthropic’s family of AI models and assistants for tasks such as writing, analysis and coding.

Lean

Lean is a programming language and proof assistant used to express mathematical statements and check formal proofs.

Editorial provenance

Version 1. Publication ID: 8bfd2c63-dd26-d2ab-c2f9-334719a262e7.

Prepared by openai/gpt-6-luna; grounding review by anthropic/claude-opus-5.5.

Source check: passed by anthropic/claude-opus-5.5 on 2026-10-29T17:09:09.675Z. This checks fidelity to the source, not whether its claims are independently true.

Editorial selection: accept; reasons: bridge_admitted_by_model, second_opinion_categorical.

Observation: d4fea46f337b87c238e5a0f0350f0a02ca2392422f778722079853bd95260ff1. Source text hash: cad4314b74c030e1bc0f85dc9da9c758da69ed3a1579bd494f1e6640af1c6150.

passed · 2026-10-29T17:09:09.675Z · anthropic/claude-opus-5.5

Publication: 8bfd2c63-dd26-d2ab-c2f9-334719a262e7

attribution — pass: The reading attributes the proof claims, the 11-day timeline, Prove2Me's role and the outlook on formalization to Anthropic. Buzzard's quoted characterization is correctly attributed to him. The source describes Buzzard as having kicked off the community formalization effort, which reasonably supports 'whose team had worked on formalizing the theorem.'

a multi-year community effort kicked off in 2024 by Kevin Buzzard at Imperial College London to complete the formalization using the Lean proof assistant

summary — pass: The summary leads with the actor (Anthropic and Claude) and the development. It explains Lean and formalization in plain language and notes the role of Prove2Me. It conveys the argued significance, reducing the review burden, along with the caveat that formal proofs should not replace human exposition. The question is a genuine open issue that follows from the source. Fermat's Last Theorem is not restated, but it is widely known and its absence does not block understanding.

Proof assistants like Lean verify the logic of a proof algorithmically, demonstrating its correctness beyond a doubt.

date — pass: evidence.originalPublished is the instant 2026-09-04, and the reading states September 4, 2026, which matches the UTC day. Nothing is presented as happening after that date; the dated passage in the source, the Aug-18 thinking excerpt, comes earlier. The page was edited on 2026-10-14, but the claims rely on core body text that is plainly part of the article, and the limitations disclose the edit. The source is a publisher page, so no archive disclosure is required. Forecasts are framed as Anthropic's arguments.

The FLT root reads PROVED on prove2me at 02:00:57Z Aug-18 (10:00:57pm ET Aug-17).

support — pass: Each claim matches the source text. The source supports the 11-day, largely autonomous end-to-end proof, the 13 million lines of Lean, and Prove2Me's coordination and reuse functions (the DAG and natural-language descriptions). It also supports the argument about reducing the review burden and the caveat about human exposition. The implied benefits are kept as Anthropic's argument, and the open questions are labeled as unknown.

Enabling search and reuse by maintaining a natural-language description of each theorem statement, resulting in a simpler proof path.

media — pass: No media candidates were supplied, and the reading selects none, which is appropriate.

Challenge or correct this reading

Send a correction or an opposing source with this reading’s URL. Disagreement and uncertainty can remain unresolved.