OpenAI’s September 8, 2026 announcement that an internal frontier model directed roughly 10,000 autonomous agents to produce a Lean-verified proof of finite-time singularity in the forced 3D Navier-Stokes equations set the benchmark for AI-driven Millennium Prize claims. The result addressed one of the Clay formulation’s allowable statements but triggered scrutiny over timing relative to concurrent human-AI work by NYU’s Tristan Buckmaster and Anthropic’s Levent Alpöge on related Euler equations. In late October, OpenAI released further AI-generated progress on the Yang-Mills existence problem; a November 10 review by a panel led by Nobel laureate David Gross confirmed a partial formalization in Lean while noting the mass-gap requirement remains unmet. Clay still lists both problems as unresolved pending publication, two-year community validation, and full compliance with prize rules. Traders are watching for any lab’s next verified claim on remaining problems such as P vs NP or the Hodge conjecture, alongside continued independent verification of OpenAI’s submissions and potential competitive announcements from Anthropic or other labs.