Recent OpenAI announcements of an internal frontier model generating solutions or advances on over 370 previously open problems—including partial results tied to the Hodge Conjecture, quasi-Riemann hypothesis, and Navier-Stokes—have elevated trader expectations for qualifying full solutions by year-end. These releases, following September claims on Millennium Prize problems and similar efforts from doubleAI on graph theory and AlphaProof Nexus on Erdős problems, highlight accelerating large language model capabilities in formal proof generation and Lean verification. However, many results remain unverified or partial, sparking debate among mathematicians about attribution, completeness, and credit. Competitive pressure among OpenAI, DeepMind, and Anthropic, plus upcoming model iterations, could drive additional verified announcements before December 31.