Yang–Mills Mass Gap: What OpenAI Proved in 2026

Quick answer: The Yang–Mills Millennium Problem asks for two things: a mathematically rigorous four-dimensional quantum Yang–Mills theory, and a proof that it has a mass gap (its lightest particle has positive mass). In October 2026, OpenAI announced an AI-built proof of the existence part for pure Yang–Mills; in November 2026 a committee chaired by Nobel laureate David Gross reported that the Lean code establishes existence. The mass gap was not claimed and remains open.

Yang–Mills theory describes the strong and weak nuclear forces and is a pillar of the Standard Model of particle physics. Physicists use it with great success, yet mathematicians have never been able to define it rigorously in four dimensions, or prove its most important feature. That is the Yang–Mills existence and mass gap problem, worth 1 million dollars.

The Problem in Two Parts

PartWhat must be shownStatus (November 22, 2026)
ExistenceFor any compact simple gauge group, a quantum Yang–Mills theory exists on 4D spacetime satisfying standard axioms of quantum field theoryAI-built Lean proof for pure Yang–Mills, backed by an expert committee; not yet peer-reviewed in the usual way
Mass gapThe theory’s lightest excitation has strictly positive mass \(\Delta > 0\)Open

The Clay prize requires both parts.

What Is a Mass Gap?

Gluons, the carriers of the strong force, are massless in the equations. Yet experiments show the strong force acts only at short range, as if its particles had mass. The mass gap would explain this: the lightest states (glueballs) have positive energy, so the force cannot reach far. It is related to, but not the same as, confinement, why quarks are never seen alone. Experiments and computer simulations strongly support a mass gap; the challenge is a proof.

What Happened in 2026?

  • October 26, 2026: OpenAI announced that its internal model had produced (in natural language on September 9) an existence proof for four-dimensional pure Yang–Mills theory satisfying the Wightman axioms, built as a scaling limit of lattice gauge theory, with a Lean formalization completed on October 16. It did not claim a mass gap.
  • Expert review: OpenAI approached David Gross (Nobel Prize in Physics 2004), who formed a working group with Michael Douglas, Arthur Jaffe, Igor Klebanov and Nikita Nekrasov.
  • October 28 to November 10, 2026: the group assessed the Lean submission received October 28, announced its release on November 9 and published its assessment on November 10: after rebuild, axiom audit and statement checks, the Lean code establishes the existence of pure Yang–Mills theory under standard definitions. The submitted construction neither proves nor disproves a mass gap.

Reviewers have said they do not yet fully understand the proof itself; the assessment covers the Lean submission, not a separate informal manuscript. OpenAI has separately announced work on QCD, which the public assessment does not cover. The working group says it has no connection to the Clay Institute. Beyond the missing mass gap, which is the substantive gap, Clay’s rules also require publication, acceptance and a waiting period of at least two years.

What Remains Open

  • The positive mass gap.
  • Confinement, a related but separate goal mentioned in Clay’s problem description.
  • The area law for Wilson loops, a standard signal of confinement.
  • Independent assessment of QCD claims, and uniqueness questions.

Common Misconceptions

“AI solved the Yang–Mills Millennium Problem.” No. Only the existence half is claimed, and the prize requires the mass gap too.

“A Lean proof settles everything.” Lean checks the logic; experts still had to confirm the definitions match the standard ones, which is what the Gross group reported.

Watch: Yang–Mills and the Mass Gap in Video

If you learn better by watching, these videos are worth your time.

The Yang-Mills Problem (news) (Numberphile). A November 2026 explanation of the news and what it does and does not prove.
How Close Are We To Proving Yang-Mills and Mass Gap? (Voyager). Background on the physics and why the mass gap is so hard.

Check Your Understanding

Five quick questions to see how much of this has stuck. Pick an answer and press Submit; you will see a short explanation either way.

Answer each question and press Submit to see your result right away.

1. What two things does the Yang–Mills Millennium Problem require?
Found a bug? Let us know!
2. What did OpenAI claim in October 2026?
Found a bug? Let us know!
3. Who chaired the expert committee?
Found a bug? Let us know!
4. What does a positive mass gap mean?
Found a bug? Let us know!
5. Has the Clay prize been awarded for Yang–Mills?
Found a bug? Let us know!

Frequently Asked Questions

What is the Yang–Mills mass gap problem?

It is a Millennium Prize Problem asking for a rigorous construction of four-dimensional quantum Yang–Mills theory and a proof that its lightest particle has strictly positive mass.

Did OpenAI solve the Yang–Mills problem?

Only the existence part. In October 2026 OpenAI announced an AI-built existence proof for pure Yang–Mills, and a committee chaired by David Gross reported in November 2026 that the Lean code establishes existence. The mass gap was not claimed and remains open.

Why does the mass gap matter?

It would explain why the strong nuclear force is short-ranged and is closely tied to confinement, the fact that quarks are never seen alone.

Is the 1 million dollar prize still available?

Yes. The Clay problem requires both existence and a mass gap, and Clay has made no prize decision.

Keep Exploring

Sources and Further Reading

Related to This Article

What people say about "Yang–Mills Mass Gap: What OpenAI Proved in 2026 - Effortless Math"?

No one replied yet.

Leave a Reply