OpenAI Used Pirated Books to Train AI, Authors Allege — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- OpenAI used books from a 'sketchy Russian website' to train GPT models, according to internal knowledge documented in court filings, as part of a strategy to win the AI arms race
- Microsoft is co-defendant in the Authors Guild v. OpenAI lawsuit, accused alongside OpenAI of illegally training AI on copyrighted books without permission or payment
- Authors Guild filed explosive new documents alleging OpenAI and Microsoft acted with 'guilty knowledge' and 'reckless disregard' for authors’ rights and the future of American literature
- Mary Rasenberger, CEO of the Authors Guild, stated executives showed 'shocking disdain for writers' by choosing mass book piracy despite knowing it would destroy author livelihoods
- David Baldacci, George R.R. Martin, Jodi Picoult and other prominent authors are plaintiffs in the case, which warns AI-generated content threatens to replace human-authored works with 'AI slop'
- The Authors Guild argues OpenAI released products that would 'make people unemployed' and degrade culture, while fearing negative 'optics' if such plans became public
Why it matters: Over 18,000 working authors stand to lose control of their copyrighted works and potential income streams if AI companies are allowed to use books without licensing. The case sets a precedent for whether generative AI can legally be built on unlicensed creative content at scale.
Ask SkimNews
