● Covered by 1 source · 3 reportsMedium impact2 neutral1 positive
Anthropic's Claude AI Formalizes Fermat's Last Theorem in Lean
Key points
- Claude AI formalized Fermat's Last Theorem in Lean.
- The proof was generated autonomously over 11 days.
- It completes Freek Wiedijk's 100 formalization challenges.
- The proof relies on Lean's three standard axioms.
- The argument follows Wiles and Taylor-Wiles methods.
AI Achieves Mathematical Milestone
Anthropic's Claude AI has produced a complete computer-checked proof of Fermat's Last Theorem (FLT) using the Lean programming language. The AI model worked largely autonomously over 11 days to formalize the proof, which was then verified by both the Lean kernel and an independent Rust-based kernel, nanoda.
Completion of a 20-Year Benchmark
This formalization marks the completion of Freek Wiedijk’s list of 100 formalization challenges, a benchmark in mathematical formalization that has existed for 20 years. The achievement was officially announced by Anthropic, following its generation using an internal model and the prove2.me platform.
Mathematical Details and Verification
The proof relies solely on Lean's three standard axioms: propext, Classical.choice, and Quot.sound. The argument follows the methods of Frey, Serre, Ribet, Wiles, and Taylor-Wiles, specifically the Darmon–Diamond–Taylor exposition from 1995. The formalization includes the development of Fontaine theory and Mazur’s work on the Eisenstein ideal, which helps conclude that no Frey curve can have a point of order, making the FLT proof applicable for n > 2.
Historical Context of Fermat's Last Theorem
Fermat's Last Theorem, first conjectured around 1637 by Pierre de Fermat, states that no positive integers a, b, c satisfy aⁿ + bⁿ = cⁿ for any n > 2. The first proof, provided by Sir Andrew Wiles in 1995, was 129 pages long and required extensive verification. The concept of formalizing such proofs, converting mathematical reasoning into a computer-checkable format, was proposed by Dutch computer scientist Jan Bergstra a decade later.
Implications for Research Mathematics
This work demonstrates significant progress in AI's capability to formalize complex mathematical proofs. The ability of AI to autonomously generate and verify such intricate mathematical arguments could influence future research and verification processes in mathematics.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
Primary sources
GitHub anthropics/fermats-last-theorem GitHub leanprover/comparator GitHub ImperialCollegeLondon/FLT GitHub leanprover-community/flt-regular GitHub flyspeck/flyspeck GitHub cmc_internal/apiHow outlets covered it
Anthropic announced that one of its internal AI models, using the prove2.me platform, has formalized a complete proof of Fermat’s Last Theorem (FLT) in Lean. This achievement marks the completion of Freek Wiedijk’s list of 100 formalization challenges, a 20-year benchmark in mathematical formalization.
Fermat's Last Theorem has been completely machine-checked in Lean 4, using the Mathlib library. This proof relies only on Lean's three standard axioms and was verified by both the Lean kernel and an independent Rust-based kernel, nanoda.
Anthropic's Claude AI autonomously generated the first complete computer-checked proof of Fermat's Last Theorem in the Lean programming language over 11 days. This achievement demonstrates significant progress in AI's ability to formalize complex mathematical proofs, potentially impacting the future of mathematical verification.