The Forced Alternative Is a Filter: What a Constructed Navier-Stokes Blowup Proves, What It Does Not, and One Order-Type Demand on Its Correction Tower
A direct reading of the OpenAI forced Navier-Stokes construction of 8 September 2026. On 8 September 2026 a construction was released, under the author line OPENAI, claiming a smooth compactly supported force under which a solution of the three-dimensional incompressible Navier--Stokes equations develops unbounded velocity in finite time from rest at bounded energy, thereby resolving Millennium alternatives (C) and (D). This paper does not assert that the construction is wrong. It proves four things about what such a construction can mean and one thing it must supply. First, a reduction theorem: alternative (C) holds if and only if there exists a smooth bounded-energy field, singular in finite time, whose Navier--Stokes residual extends to a smooth decaying force. The equation therefore acts as a regularity filter on a field one is free to prescribe, and no evolution from data enters the forced alternatives at all. Second, the construction's own limit lemma places the force flat to infinite order at the singular point, so the force performs no act where the velocity diverges. Third, the constructed field is the limit of an iteration designed to make the residual flat, so the theorem is true by design conditional on convergence, and in the source-attribution register the field is occupied by its own designer and by no generator independent of the design. Fourth, in the construction's dyadic chart the small parameter is a function of the band index and the stage count within a band is band-independent, so the correction scheme has order type omega squared; the anchor such a tower requires is a modulus of moduli, uniformity of the band constants in the band index, and per-band control does not compose by itself. We state that demand as the paper's one open question to the authors and to any referee. We close with the complementary unforced criterion of an earlier paper and with three falsification criteria for the present one. METHODSole author; drafted with a large language model as scribe under the author's direction, disclosed in Appendix A with the two concessions the model forced. Self-audited in five adversarial rounds before release; the complete census of four findings earned against the paper, and one fault charged against the audit, is printed as Appendix C. A reproduction script, Appendix B and CHECK.py, verifies every executable claim in about two seconds and reports 14 of 14. Companion to DOI 10.5281/zenodo.22665831. STANDINGThis paper does not assert that the OpenAI construction is wrong, and would be unharmed by its confirmation. Not peer reviewed; no machine verification of its own claims is offered. ELSEWHERESource, script and revision history: https://github.com/1000sapients/Trisduction/tree/main/Publication%20Library/Mathematics/Analysis/Navier-Stokes%20Regularity Rights: CC BY 4.0. Mohammad F Islam, PhD, 2026.