Twenty-five recipients of the Fields Medal — the highest individual award in mathematics, typically described as the mathematical equivalent of the Nobel Prize — signed an open letter published on 11 September condemning OpenAI's conduct around its Navier-Stokes Millennium Prize claim. The signatories demand that AI labs solving famous mathematical problems provide "time for a proper writeup, isolation of new methods, and citing relevant previous work" before public announcement. The letter raises a concern beyond the immediate attribution dispute with NYU mathematician Tristan Buckmaster and Anthropic researcher Levent Alpöge (reported in Issue 200): that mathematicians' own contributions to AI training collaborations — including work shared with Codex and other OpenAI systems — may have been absorbed into training data without attribution, threatening the open-research culture that has produced the mathematical knowledge AI systems are trained on. OpenAI simultaneously withdrew its sponsorship from a Caltech mathematics conference after researchers there published public criticism of the company. The collective rebuke from 25 Fields Medalists — not a fringe group but the field's most formally recognised practitioners — constitutes an institutional position. Independent peer review of OpenAI's Navier-Stokes proof submission remains pending; no Clay Mathematics Institute determination has been announced.