Sarah Silverman and Authors Sue OpenAI and Meta Over AI Training Data
Comedian Sarah Silverman and two other authors file proposed class action lawsuits against OpenAI and Meta, accusing the tech giants of using their copyrighted books to train AI models without permission.
Comedian Sarah Silverman and authors Christopher Golden and Richard Kadrey file a pair of proposed class action lawsuits against OpenAI and Meta Platforms in a California federal court. The comedians and writers claim that both technology companies use their copyrighted books as training material for their artificial intelligence generators without asking for permission or offering compensation.
The legal complaints allege that OpenAI's ChatGPT generates highly accurate summaries of the plaintiffs' books when prompted, proving the AI system retains specific knowledge of their copyrighted works. Additionally, the authors assert that many of their books appear in the dataset that Meta admits to using to train its LLaMA language model, which the lawsuits classify as direct copyright infringement.
These lawsuits seek a permanent injunction, financial damages, and recovered profits, while also accusing the companies of violating the Digital Millennium Copyright Act and engaging in unfair competition. This legal action joins a rapidly growing wave of litigation against generative AI companies, emerging at a time when courts and lawmakers have yet to decide if scraping copyrighted material to train AI models constitutes a legal violation.