Publishers sue Meta, saying it violated copyrights in training AI with their books

Five leading publishers and a best-selling author filed a class-action lawsuit against Meta and its founder, Mark Zuckerberg, on Tuesday, alleging that the tech giant violated copyright law by training its generative artificial intelligence platform on millions of illegally pirated books and articles.

Author Scott Turow and the publishers named as plaintiffs in the lawsuit — Hachette, Macmillan, McGraw Hill, Elsevier, and Cengage — allege that Meta and Zuckerberg “illegally torrented millions of copyrighted books and journal articles” and “downloaded unauthorized web scrapes of virtually the entire internet.” Meta then copied the stolen material many times over, the plaintiffs allege, to train the AI system — effectively engaging “in one of the most massive infringements of copyrighted material in history.”

The case, filed in the U.S. District Court for the Southern District of New York, is the latest in a string of lawsuits brought by publishers, authors, artists, photographers, and news outlets aimed at forcing tech companies to compensate them for using their works to train their AI models. The plaintiffs argue in the lawsuit that the AI model’s ability to quickly produce knockoffs and summaries of copyrighted books threatens the livelihoods of publishers and authors.

A Meta spokesperson said in a statement that the company would “fight this lawsuit aggressively.”

“AI is powering transformative innovations, productivity and creativity for individuals and companies, and courts have rightly found that training AI on copyrighted material can qualify as fair use,” the spokesperson said.

The publishers’ complaint states Meta distributed millions of copyrighted works without authorization and without compensating authors or publishers, contending that Zuckerberg “personally authorized and actively encouraged the infringement.” They also say that Meta removed copyright notices and copyright management information from the works used to train the AI model, known as Llama.

Tuesday’s lawsuit mirrors claims made against the company Anthropic, which was sued by authors over its use of their work to train its AI tools, such as its popular chatbot Claude.

The Washington Post reported earlier this year that the AI company spent tens of millions of dollars to acquire and slice the spines off millions of books, before scanning their pages to feed more knowledge into the models behind its products. The details of the effort, known as Project Panama, were outlined in documents unsealed in a now-settled case brought by authors against Anthropic.

Court records in the settled case suggested that tech companies didn’t see it as practical to gain direct permission from publishers and authors to use their work. Instead, Anthropic, Meta, and other companies found ways to acquire books in bulk without the authors’ knowledge, court filings allege, including by downloading pirated copies.