OpenAI’s September 2026 solution to the Navier-Stokes equations using swarms of AI agents, followed by its early October release of hundreds of additional proofs including a claimed resolution of the rational Hodge conjecture for CM abelian varieties, has shifted trader focus toward the Hodge Conjecture as the next full Millennium Prize breakthrough. These results, many Lean-verified and spanning algebraic geometry and number theory, highlight how large language models combined with agentic workflows can tackle abstract problems that align with current training strengths in formal reasoning. Birch and Swinnerton-Dyer sits close behind due to related elliptic curve progress, while P versus NP and the Riemann Hypothesis trail because their core complexity and analytic demands remain farther from demonstrated AI capabilities. With “no solution by end-2027” nearly tied for the lead, markets reflect uncertainty over verification timelines, potential partial advances versus full Clay Institute acceptance, and competition among labs like DeepMind and Anthropic.