OpenAI reports a proof of Barnette’s conjecture

OpenAI reports that every finite simple cubic bipartite planar graph that is 3-vertex-connected has a Hamiltonian cycle.

Sources for this record

What was already known

Barnette’s conjecture had remained open since 1969. Earlier proofs covered restricted face sizes, including the quadrilateral–hexagonal case and the faces-at-most-eight case.

AI involvement AI-led

OpenAI credits its internal frontier model with producing the mathematics release and lists OpenAI as author of this manuscript. The source does not establish a more precise account of human input for this proof.

OpenAI’s release states that its internal frontier model produced the results, and OpenAI is credited as author of this manuscript. The individual human contribution is not detailed enough to establish autonomy.

Importance Major

Resolves Barnette’s longstanding Hamiltonian-cycle conjecture for the full stated graph class.

Limitations

The paper and Lean development are OpenAI-authored reports. The scope note says the Lean theorem matches the full graph statement, while the repository records its formalization review as unchecked. No named human mathematical explanation or independent proof review is documented here.

Technical verification Formal (Lean 4)

OpenAI’s pinned scope note says the Lean development proves the full selected Barnette statement, and the named main theorem and Comparator configuration are in the repository. This is source-reported machine checking; the repository review status is unchecked and no independent run is recorded here.

Formal checking

The repository scope note says its Lean development proves the full selected theorem; the matching main declaration is in the linked source.

Checker
OpenAI-reported Lean 4 / Comparator
Scope
Every finite simple cubic bipartite planar 3-vertex-connected graph has a Hamiltonian cycle.
Basis
Reported by the linked checker
Evidence source

Checking a formal statement does not establish that it matches the original problem.

Human understanding Few

Few is the default when evidence for a higher level is insufficient. The paper explains its proof, but OpenAI is the named author and no named person is documented as reconstructing the central paired-state argument. Few is the evidence default, not a report that people cannot understand it.

Correction and publication history

2026-10-18 · publication

Primary manuscript, prior-work and pinned formalization review of one OpenAI-authored contribution; historical non-notifying inclusion.

Read this published revision