{"slug": "openai-pulls-three-ai-generated-math-papers-one-day-after-release", "title": "OpenAI pulls three AI-generated math papers one day after release", "summary": "OpenAI withdrew three of the 722 AI-generated mathematics papers it released on October 6, 2026, after finding a sign error in one algebraic geometry proof that undercut key arguments in two dependent manuscripts, leaving a corrected catalog of 719 papers. Roughly 300 of the 719 results had been formally verified in the Lean proof assistant, about 42%, and OpenAI said the mathematical assertions themselves have not been refuted. The release, produced by an unnamed proprietary model that spent about three hours per result, was informed by the Advisory Group on Mathematics and Artificial Intelligence (AGMAI).", "body_md": "OpenAI official logo (public domain, Wikimedia Commons) — CryptoBriefing brand treatment\n\n# OpenAI pulls three AI-generated math papers one day after release\n\nA sign error in one algebraic geometry proof forced OpenAI to withdraw papers from its 722-manuscript math release, reviving questions about verification\n\n[OpenAI](https://cryptobriefing.com/markets/openai/) released 722 AI-generated mathematics papers on October 6, 2026. By October 7, three of them were gone.\n\nThe company withdrew the papers after finding a sign error in one proof. That single mistake undercut key arguments in two related manuscripts that depended on it.\n\n## What OpenAI released, and what it pulled\n\nThe original drop landed on OpenAI’s GitHub repository. It contained 722 papers grouped into 372 result families, all produced by an unnamed proprietary model.\n\nThe manuscripts aimed at around 4,000 unresolved problems in mathematics. They claimed advances on numerous longstanding conjectures, including progress on several Millennium Prize Problems.\n\nThe model reportedly spent approximately three hours on average per result.\n\nThe flawed proof sat in algebraic geometry. A sign error means a plus where a minus should be, or the reverse. In a proof, one flipped sign can turn a valid chain of logic into one that proves nothing at all, and every argument built on top of it falls with it.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nAfter the withdrawals and revisions, the corrected catalog now stands at 719 papers. OpenAI emphasized that the mathematical assertions themselves have not been refuted.\n\n## The verification problem\n\nAfter the initial release, 300 of the 719 results had been formally verified in Lean, a proof assistant that translates a proof into formal code and checks it step by step. That works out to about 42%.\n\nCritics also pointed to the proprietary nature of the model. Outside researchers cannot inspect it or reproduce its process.\n\nThe release was informed by the Advisory Group on Mathematics and Artificial Intelligence, known as AGMAI.\n\n## A pattern of friction with mathematicians\n\nThis is not OpenAI’s first clash with the math community. The company previously withdrew a $1M sponsorship from a Caltech AI math contest after criticism from 771 mathematicians.\n\n## What this means\n\nThe formally verified results carry a strong guarantee. The remaining majority carry the same kind of risk that produced the sign error.\n\nOpenAI’s insistence that the claims remain unrefuted is technically meaningful but limited. An unrefuted claim with a broken proof is, mathematically speaking, back to being a conjecture.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our\n\n[Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/openai-pulls-three-ai-generated-math-papers-one-day-after-release", "canonical_source": "https://cryptobriefing.com/openai-withdraws-ai-generated-math-papers/", "published_at": "2026-10-09 00:52:57+00:00", "updated_at": "2026-10-09 01:17:22.399740+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-research", "large-language-models", "ai-safety"], "entities": ["OpenAI", "Lean", "Advisory Group on Mathematics and Artificial Intelligence", "AGMAI", "Caltech", "GitHub", "Diego Almada Lopez", "CryptoBriefing"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/openai-pulls-three-ai-generated-math-papers-one-day-after-release", "markdown": "https://wpnews.pro/news/openai-pulls-three-ai-generated-math-papers-one-day-after-release.md", "text": "https://wpnews.pro/news/openai-pulls-three-ai-generated-math-papers-one-day-after-release.txt", "jsonld": "https://wpnews.pro/news/openai-pulls-three-ai-generated-math-papers-one-day-after-release.jsonld"}}