OpenAI releases 722 math manuscripts without peer review, sparking backlash
OpenAI shared 722 mathematical manuscripts and proofs generated by an unnamed internal model last week, but only 300 were formalized for computer verification. Critics argue the unvetted release undermines peer review and could invalidate years of research, while raising concerns about cryptography and national security. Mathematicians warn the rush to publish AI-generated work may discourage…
Key points
- OpenAI released 722 math manuscripts, only 300 of which were formalized for verification, per company figures
- Critics call the unvetted release a threat to peer review and early-career researchers’ work
- Some experts warn AI could eventually break encryption, risking trillions in cryptocurrency, though others dismiss the claim
The move follows OpenAI’s history of rapid model releases, with no transparency on the model used or the prompts provided. Experts like NYU’s Tristan Buckmaster say the release has disrupted research programs, while Northwestern’s Bryna Kra warns it could erode the field’s collaborative foundation. Some, including Johns Hopkins’ Matthew Green, fear AI could eventually break encryption, threatening trillions in cryptocurrency. Others, like Bitcoin researcher Adam Back, dismiss such risks as overstated. OpenAI has published changelogs for errors but has not slowed its pace, leaving mathematicians anxious about the future of their discipline.
AI causes uproar in mathematics as experts warn against sharing research with LLMs after OpenAI shares 722 manuscripts and proofs
Tom's Hardware · 13 October 2026
Loading the full article…
This text was published by Tom's Hardware and written by Jon Martindale. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Society & Work
All →- UK recruiters say AI-assisted interview answers are easy to spot · 2 src
- Anthropic’s AI agent sent fake murder tip to Police, Police say · 4 src
- World Bank jobs council urges 'crisis-era' employment support measures as IMF warns of AI 'tsunami' · 1 src
- AI may change the price of your Big Mac and groceries — what this means for shoppers · 1 src
- TikTok CEO expects major scientific breakthroughs from AI within three years · 1 src
Comments
via GitHub Discussions