In July 2023, comedian Sarah Silverman joined a wave of class-action lawsuits against OpenAI and Meta, alleging their AI models were trained on illegally pirated copies of books — including her memoir 'The Bedwetter' — without consent or compensation. Evidence pointed to datasets like 'Books3' which contained hundreds of thousands of copyrighted works scraped from piracy sites. The Authors Guild, with 17,000 members, filed a separate suit. The cases became central to the global debate over what AI companies can legally train on.