OpenAI’s early October 2026 release of 722 papers covering 372 families of previously unsolved mathematical results—generated by an unreleased internal model—has become the dominant driver of sentiment on whether Epoch AI will soon classify another FrontierMath Open Problems entry as solved. These outputs include Lean formalizations and claimed advances tied to Millennium Prize Problems such as Navier-Stokes and aspects of the Riemann hypothesis, following OpenAI’s September announcement on Navier-Stokes. Traders are weighing the speed and scale of OpenAI’s multi-agent approach against verification standards, as Epoch AI’s official classification—not third-party claims—determines resolution. Competitive efforts from Anthropic on formalization milestones and smaller labs like doubleAI on specific graph-theory improvements add context, while upcoming Epoch AI updates to its open-problems tracker remain the key near-term catalyst.