Authors accuse OpenAI of training GPT on pirated books, seek summary judgment

A group of authors filed a motion for summary judgment in New York accusing OpenAI of training its GPT models on pirated LibGen copies, concealing the piracy and building the business on mass piracy, and asking the court to find liability and that such copying isn’t fair use; the filing covers 194 titles and also argues Microsoft is vicariously liable, while OpenAI counters with a cross-motion asserting fair use as part of broader lawsuits against OpenAI and Microsoft.