Microsoft is facing a lawsuit in New York federal court from a group of authors who say the company used their books without permission to train its artificial intelligence model called Megatron.
The lawsuit was filed on Tuesday by writers including Kai Bird, Jia Tolentino, and Daniel Okrent. They claim Microsoft used pirated digital versions of nearly 200,000 books to train the AI system, which generates text responses to user prompts.
The authors are seeking a court order to stop the alleged infringement and damages of up to $150,000 for each work.
The complaint says Microsoft built its model using copyrighted books without authorization. It also says the AI was trained to mimic the style, themes, and voice of the original texts. The writers say this violates copyright law and harms their rights as creators.
The lawsuit comes one day after a federal judge in California ruled that Anthropic made fair use of copyrighted materials in training AI models, but could still be liable for using pirated books. That case was the first court decision in the U.S. to address the use of copyrighted content for generative AI training.
Microsoft did not respond to a request for comment. A lawyer for the authors also declined to comment.
Several companies, including Meta, Anthropic, and OpenAI, have been sued in recent months over similar issues. The companies argue that using copyrighted works for AI training is fair use and helps create new tools without copying the original content.
