Updated November 12, 2026.
A group reviewing OpenAI’s October Yang-Mills submission says the formal construction passes its checks — an important result, but not a complete solution to the Millennium Prize Problem.
What was verified?
The assessment says the submission constructs a relativistic quantum field theory satisfying the Wightman axioms as a scaling limit of four-dimensional Wilson lattice gauge theory for any simple compact gauge group.
The formal work was checked with Lean, allowing the mathematical steps to be mechanically verified rather than accepted only as prose.
Did OpenAI solve the entire Yang-Mills Millennium Prize Problem?
No. The review group explicitly says the work does not prove the mass gap, and does not establish uniqueness of the infinite-volume limit.
That distinction matters because the Clay Millennium problem includes the mass-gap requirement. The current result addresses the existence/construction side, not the full prize problem.
Why does this matter for AI?
The result is another example of frontier AI systems contributing to formal mathematics and theoretical physics, while also showing why independent verification is essential for highly technical AI-generated work.
