Robotics & Physical AIOct 15, 2026
OpenAI withdrew three of its 722 maths preprints a day after release, and revised 14 others
Retraction Watch reported on 8 October 2026 that OpenAI, which posted 722 preprints to GitHub on 6 October describing progress on 372 maths problems by an internal model, withdrew three of them on 7 October. A sign error invalidated an argument in one manuscript and a construction used by two papers that depended on it. OpenAI also revised 14 other manuscripts with proof repairs, corrected statements, clearer hypotheses and dependencies, and one fixed citation. Cornell mathematician Alex Townsend said the cascade was not surprising and expects more errors to be found.
What it means Large AI-generated proof dumps need the same checking as human work. Errors that spread across dependent papers show why the machine-checkable formal versions are what counts.
Where it came from Retraction Watch