Giants such as Meta and Microsoft have been sued by several American writers for using pirated books to train AI models.

CTOnews.com October 20, Bloomberg local time reported Wednesday, a number of U.S. writers recently filed a lawsuit in federal court in New York, accusing Meta, Microsoft and other technology giants of using their work to train AI models without permission.

Pixabay, a group of writers, filed a proposed class action copyright lawsuit on Tuesday alleging that Meta and Microsoft used the controversial "Books3" dataset to train their large models on how to respond to human cues and instructions.

CTOnews.com Note: Authors groups claim that the "Books3" dataset contains thousands of pirated books.

At the same time, AI research firm Eleuther AI has also received allegations that the company is suspected of providing technology companies with data sets for training large models, including "Books3."

"Books3" contains text content from thousands of books taken from "shadow libraries," which the authors claim were misused as part of a dataset to train Llama 2, a large model developed by Meta and Microsoft. Because the companies in question had "obtained great value illegally" from the authors 'books, the authors' group had filed a claim for damages and an injunction prohibiting the misuse of their works.

CTOnews.com previously reported that 17 writers, including Game of Thrones authors, filed a lawsuit against OpenAI late last month accusing the company's AI program ChatGPT of "systematically stealing" their copyrighted works on a large scale, one of a series of recent legal actions by writers against AI infringement.

Related Reading:

17 authors sue OpenAI, claiming its AI program ChatGPT infringes copyright