Recent announcements from OpenAI have driven strong trader sentiment, particularly its September 2026 resolution of the Navier-Stokes equations—one of the Clay Millennium Prize Problems—via a multi-agent system running an internal frontier large language model, followed by an October release of 722 papers addressing hundreds of open problems across algebra, number theory, and topology. These results, many verified in Lean, highlight rapid progress in AI agents that combine reasoning, search, and formal proof checking to tackle research-level questions beyond IMO benchmarks. Competitive dynamics feature parallel advances from DeepMind’s AlphaProof Nexus, which resolved multiple Erdős problems and OEIS conjectures, and Anthropic models achieving counterexamples and partial results on conjectures like the Jacobian and Riemann zeta zeros. Key upcoming catalysts include mathematician reviews of the batch releases, potential Clay Institute recognition, and further model updates that could clarify which additional open problems reach formal solutions by year-end.