Axiom Math Verifies 15,000-Page 'Enormous Theorem' With AI
Axiom Math's AI swarm formally verified the 15,000-page CFSG proof and the 7,500-page Stacks project, producing 43 million and 16 million lines of code in a push to digitize mathematics.

Why it matters
- Axiom Math's AI swarm verified the 15,000-page CFSG proof on 25 October, producing 43 million lines of code.
- Nine days later, the company announced the automatic formalization of the 7,500-page Stacks project, started in 2005 by Aise Johan de Jong.
- The CFSG proof spans more than 500 journal articles by around 100 authors over a half-century.
- The team broke the CFSG into about 100 independently verifiable milestones, according to lead Niels Voss.
- The formal proof found and cleared up several mistakes in the original human proof, according to Colva Roney-Dougal of the University of St Andrews.
Axiom Math's AI swarm completed the automatic verification of the 15,000-page proof of the Classification of Finite Simple Groups (CFSG) on Sunday, 25 October — a result widely considered the longest mathematical proof ever written by humans, now formally verified by machines.
The company's Founding Mathematician, Ken Ono, son of a mathematician who grew up among architects of the original proof, watched the final moments as the system closed the verification. "My wife was with me as I watched the team countdown to completing the verification," he recalled. "God, I cannot tell you how emotional I was."
Nine days after the CFSG announcement, Axiom Math revealed it had also automatically formalized the 7,500-page Stacks project, a massive open-source mathematical reference work on algebraic geometry started in 2005 by Columbia University mathematician Aise Johan de Jong. Together, the two results represent a major AI engineering achievement and lay groundwork for digitizing all of mathematics.
The stakes are significant for both mathematics and AI research. Formal, machine-checkable mathematics gives researchers a verified foundation on which to build new results, preserves knowledge as expertise dwindles, and provides AI systems with reliable mathematical libraries. Group theory underpins fields from cryptography to physics, and the Stacks project covers the algebraic geometry vital to modern string theory and number theory.
What Did the CFSG Verification Require?
The CFSG concerns finite simple groups: sets of elements combined with an operation, such as addition or multiplication, that follow specific rules, cannot be reduced further, and contain a finite number of elements. All the permutations of a deck of cards form one such group. Évariste Galois laid the foundations of group theory almost 200 years ago. Since then, mathematicians have classified finite simple groups into 18 families, plus 26 sporadic simple groups that belong to none of them. The CFSG proof confirms that this list is complete.
That confirmation came at a cost. The proof draws on more than 500 journal articles written by around 100 authors over a half-century — a body of knowledge no individual mathematician can fully absorb. Translating this interconnected web into machine-readable code posed engineering challenges beyond those Axiom Math overcame in earlier formalizations this year.
Niels Voss, who led the CFSG formalization at Axiom Math, identified the proof's sheer length as one of the hardest problems. The proof contains multiple theorems, lemmas, and definitions, none of which can be checked until everything has been made machine-readable.
The team developed a workaround. "We processed all the literature from the CFSG, and traced the network structure of the informal proof before we were able to start working on it," Voss said. "We then broke up the statement of the CFSG into about 100 milestones — landmark results along the way that can be independently verified — which we used to guide the formalization and stay on track."
The final CFSG formalization runs to 43 million lines of code — roughly on a par with the Linux kernel.
How Did Other Mathematicians React?
Colva Roney-Dougal, a group theorist at the University of St Andrews who was not involved in the work, had long considered the CFSG a formalization impossibility. "I held it up as an example of something that was going to be incredibly hard to formalize because I had thought that the amount of computing it was going to require was just unattainable," she said. "So when I heard the announcement, I was really, really surprised."
Roney-Dougal praised the formal proof for finding and clearing up several mistakes in the original human proof. She called the result critical to formalizing any modern research in group theory, including her own work on how simple finite groups behave.
She also raised a concern. Because the verification ran automatically, few mathematicians can now extract information from the proof to share with the community. A new generation of mathematicians, she said, may have "missed a chance to gain knowledge from the doing of it."
What Made the Stacks Project Different?
Where the CFSG is one very long proof, the Stacks project is a very wide body of connected theory. De Jong started it in 2005 as a collaborative textbook and reference covering highly abstract algebraic geometry and algebraic stacks — geometric spaces central to string theory and number theory. The final Stacks formalization stands at 16 million lines of code, but its structure of thousands of interlocking definitions, constructions, and lemmas demanded a different approach.
Vasily Ilin, who led the Stacks project at Axiom Math, said the challenge centered on achieving faithfulness and coherence across thousands of interlocking definitions and theorems.
The team designed a hierarchical system of AI agents with specialized responsibilities working in parallel. The agents kept every definition aligned across the library while checking faithfulness through multi-agent audits, cross-checks against sources, and hard tests. This ensured that every element and task aligned across the entire project.
The resulting formal library is efficient, general, reusable, and maintainable, and it underpins much of modern mathematics. Anyone — or any bot — can now access the formal Stacks library and build new formal mathematics upon it.
Why Do These Two Results Matter Together?
The CFSG and Stacks formalizations demonstrate that formalization can scale along both axes mathematics requires. "They show the two ways formalization has to scale: one very long proof, and one very wide body of connected theory," Ono said.
Ono and Voss see several advantages in a future where formal mathematics becomes central:
- Rubber stamping theorems and proofs
- Preserving knowledge as expertise dwindles over time
- Building a shared library as a basis for creating new mathematics
Both remain optimistic about humanity's role once all of mathematics is formalized. They view formalization as a launchpad for humans to use and understand mathematics in new ways rather than a replacement for mathematical thought.
Ono framed the ambition as an opening, not an ending. "Aren't we excited to think more broadly about what the future of math can be?" he asked. "I'm excited by that. I'm really excited."
With the longest human proof and one of the widest collaborative libraries in mathematics now machine-verified, the path to formalizing all of mathematics has its two hardest scaling problems solved in a single month.
Original: axiommath.ai
More from Rebecca Stone
Show full bio
Correspondent covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
728 articles
Related articles
- AI System Verifies the 246 Theorem, a Milestone in Prime Number Research
- AI Brings Formal Verification to Real-World Software Code
- Google DeepMind's AlphaProof Nexus solved dozens of open math problems
- OpenAI and Anthropic Claim AI Models Refuted the Exponential Time Hypothesis
- OpenAI says AI models refuted the Exponential Time Hypothesis