FrontierSeptember 10, 2026via The Verge AI
Mathematicians want proof OpenAI didn’t use their work
Why it matters
Training data provenance and ethics are becoming a flashpoint in the frontier labs. If OpenAI used unpublished researcher work without disclosure, it raises questions about the legitimacy of capability claims and sets precedent for how labs must handle academic collaboration.
Key signals
- Second mathematician (Andreas Thom) publicly challenges OpenAI over training data sources
- Accusation: unpublished work may have contributed to ChatGPT's mathematical capabilities
- Broader row over OpenAI's transparency regarding training data origins
- Ethical and 'dishonest' behavior alleged by researcher community
- Follows earlier dispute over similar concerns
- Relates to OpenAI's 'increasingly impressive' mathematical discovery announcements
The hook
OpenAI faces a second accusation: did it train on unpublished mathematics without permission? The transparency row is heating up.
Another researcher is challenging OpenAI about the data driving its increasingly impressive array of mathematical discoveries. Just days after a bitter row erupted over whether the company's models benefited from unpublished work, a second mathematician has come forward accusing the AI giant of unet…