OpenAI's research division announced on 21 September that its latest reasoning systems had produced verified solutions to the Navier-Stokes existence and smoothness problem — one of the seven Clay Mathematics Institute Millennium Prize Problems, with a $1 million prize attached — alongside solutions to more than 100 previously unsolved mathematical conjectures. Princeton University confirmed the formation of a 12-person independent advisory group to conduct external verification of the Navier-Stokes proof.
The Navier-Stokes problem asks whether smooth, physically reasonable solutions always exist for the governing equations of fluid dynamics in three dimensions, or whether singularities — breakdowns in regularity — can emerge in finite time. It has been on the Millennium Prize list since 2000 and resisted resolution despite decades of sustained work from the field's leading mathematicians.
OpenAI published the full proof, derivation steps, and verification code, inviting external scrutiny. Mathematicians quoted in specialist publications noted the proof's reliance on a novel regularisation technique without precedent in the existing literature — suggesting the model developed an original mathematical approach rather than recombining known methods. The announcement establishes that AI reasoning systems can now produce original theoretical contributions at the frontier of formal mathematics — a capability most roadmap assessments published before 2026 placed further out.