OpenAI Accused of Using Paywalled Books for AI Training

CosmicTaco · 2025-04-02T11:16:55.997978+05:30

- A new paper by the AI Disclosures Project claims OpenAI used non-public O'Reilly books to train its GPT-4o model. - The paper suggests GPT-4o shows strong recognition of paywalled content compared to earlier models like GPT-3.5 Turbo. - Co-authors used the DE-COP method to detect potential copyrighted material in the models' training data. - OpenAI faces criticism for its data practices, though it does have some licensing agreements and opt-out mechanisms. - OpenAI did not respond to the allegations and continues to seek high-quality training data. Source: [TechCrunch](https://techcrunch.com/2025/04/01/researchers-suggest-openai-trained-ai-models-on-paywalled-oreilly-books/)

A new paper by the AI Disclosures Project claims OpenAI used non-public O'Reilly books to train its GPT-4o model.
The paper suggests GPT-4o shows strong recognition of paywalled content compared to earlier models like GPT-3.5 Turbo.
Co-authors used the DE-COP method to detect potential copyrighted material in the models' training data.
OpenAI faces criticism for its data practices, though it does have some licensing agreements and opt-out mechanisms.
OpenAI did not respond to the allegations and continues to seek high-quality training data.

Source: TechCrunch