OpenAI’s early October 2026 release of 722 papers from an unreleased internal frontier model—claiming solutions or substantial progress on over 370 longstanding problems, including partial advances on Millennium Prize items like the Navier-Stokes equations and quasi-Riemann hypothesis—has driven recent trader sentiment. Many proofs received Lean formal verification, highlighting rapid gains in large language model reasoning, multi-agent search, and automated theorem proving, though mathematicians continue scrutinizing novelty, exposition, and correctness amid ongoing reviews. Parallel efforts from DeepMind’s AlphaProof Nexus on Erdős problems and other labs underscore competitive dynamics in AI math capabilities. Key near-term catalysts include further verification outcomes, advisory group standards for AI-generated results, and any full resolutions announced before year-end that could shift probabilities on targets like the Hodge Conjecture or Ramsey numbers.