OpenAI’s September 2026 announcement of a Lean-formalized finite-time singularity for the forced Navier-Stokes equations (Clay statements C and D) continues to anchor trader sentiment at 94% against retraction. The result emerged from an unreleased multi-agent system running roughly 10,000 agents for 88 hours plus verification, distinguishing it from the three papers OpenAI quickly withdrew in early October over a sign error. Ongoing disputes center on credit, data provenance, and whether the forced construction addresses the deeper unforced problem, yet no independent refutation or company acknowledgment of a fatal flaw has surfaced. Realistic reversal risks remain if peer review uncovers an unfixable error before year-end or if OpenAI’s internal model produces contradictory follow-up work.