The Lean and Mathlib communities issued a joint statement on the 21st, certifying the Lean proof of the Riemann Hypothesis collaborated by Anthropic mathematicians Alpöge and Claude. Both the proposition and the proof are correct. (Preliminary summary: Was the Riemann Hypothesis broken by Claude? Mathematician Levent Alpöge claimed to have completed the proof, and the huge argument exceeded 250,000 pages) (Background supplement: It was just reported that Claude cracked the Riemann Hypothesis, and a purely human version is also coming! The president of the American Mathematical Society posted a 940-page proof draft, emphasizing that "it does not use AI at all") The proof of the Riemann Hypothesis has passed the computer verification level! On November 21, Lean and the Mathlib community issued a joint statement, expressing "full confidence" that Anthropic mathematician Levent Alpöge's Riemann Hypothesis was proven correct, and certified that both the proposition and the proof were correct. Lean’s official X account posts simultaneously. Lean is a programming language that allows computers to check mathematical proofs step by step. Mathlib is a mathematical library maintained by the Lean community and reviewed by experts. This statement is also endorsed by the Lean non-profit organizations Lean FRO and Mathlib Initiative. Alpöge published the proof of the 17 million lines of code on November 20, and the community completed the certification about a day later. According to Alpöge's explanation, he "cooperated with Claude" to find the proof, and the direction was guided by him. In the final stage, Claude was significantly ahead of him, and the formalization was completed by him and Claude. The joint statement itself made no mention of Claude. The Lean FRO and Mathlib Initiative have released a joint statement on the Lean verification of the Riemann Hypothesis. On November 20, Levent Alpöge released a proof of the Riemann Hypothesis as a Lean library of 17 million lines of code. Over the past day, the Lean and Mathlib… pic.twitter.com/pYR61ZaREc — Lean (@leanprover) November 21, 2026 It has been tested on a computer, why is it trustworthy? The computer only checks the reasoning, regardless of whether you are proving the correct question, so the focus of verification is to block the room for cheating. Kevin Buzzard, a mathematician at Imperial College London, listed four questions to be confirmed on his blog, and the joint statement said that all of them were met. The first level is the proposition. The definition of Riemann Hypothesis is taken from Mathlib's built-in Riemann Hypothesis. Lean has carefully reviewed it with number theory experts. It is not the version written by Alpöge himself. The second level is compilation. Dozens of people recompiled on their own machines and got the same certificate file. MIT mathematician Andrew Sutherland's verification page records that all 175,187 modules were compiled successfully with zero errors. These checks were performed using Claude Code. The third level prevents stealing. The community added an additional file that only quoted Mathlib and specifically wrote out the Riemann Hypothesis propositions. It then used two independent tools, the comparison mode of comparator and con-leche, to confirm that the theorem proved by Alpöge was completely consistent with Mathlib's Riemann Hypothesis. The fourth level of prevention is to prove the vulnerability of Lean itself. The exported certificate files are submitted to multiple checkers, including Lean official core, con-leche, con-ron, and third-party checkers written in different programming languages, all of which are accepted. The two checkers, con-leche and con-ron, have also been formally proven, ensuring that they will not accept wrong proofs. The check is run on the exported file that does not execute the program, proving that even if there is a malicious program hidden in the source code, it cannot be encountered. Is this a solution? At the level of the machine, Sutherland told Scientific American that the Lean proof has been compiled and independently rechecked by the Lean core, using only three standard axioms of mathematics, and the chance of a problem is "approximately zero." What is lacking is human understanding. Alpöge only makes public the Lean code and does not write a human-readable version of the paper. He himself said that he has not fully grasped the argument yet and plans to compile a version that adults can understand. At the end of the joint statement, we also look forward to the mathematical community's help in explaining the new ideas in the proof. Ranges also have boundaries. The joint statement endorses only the Riemann Hypothesis. Alpöge claims that the method can be extended to the generalized Riemann Hypothesis, but the MIT verification page only checks the Dirichlet L function part, and other types of L functions are not within the scope. The page also indicates that there are two more checks running. In summary: The machine has already nodded, and it will take how long it will take for humans to understand. Related reports Claude challenges the Riemann Hypothesis that humans have not solved for 167 years. This is the current progress. Can you still believe the AI proof if you can’t understand it? Mathematician: It’s like a new drug. OpenAI throws out 722 mathematics papers at once. Some scholars’ years of hard work turned to zero overnight. OpenAI claims to have solved the mathematical problem of the century, but it fell into “misappropriation of other people’s undisclosed research.” Anthropic’s IPO price is higher than expected at US$83, and the raised amount of US$81.3 billion surpasses SpaceX. Computer verification of the Riemann Hypothesis passed! Lean community joint statement: "Have complete confidence" in the correctness> This article was first published in DongZu BlockTempo "DongZu DongTen - the most influential blockchain news media".
Riemann Hypothesis passed computer verification! Lean community joint statement: "complete confidence" in correctness
The Lean and Mathlib communities issued a joint statement on the 21st, certifying that Anthropic mathematician Alpöge and […] Riemann Hypothesis passed the computer verification! Lean community joint statement: "Have complete confidence" in the correctness> This article was first published in DongZu BlockTempo "DongZu DongTen - the most influential blockchain news media".

This article may contain an automated translation. The original source is linked for verification.



