Elsevier, Cengage, Hachette, Macmillan and McGraw-Hill sued Meta Platforms in Manhattan federal court on Tuesday, accusing the tech giant of misusing its books and journal articles to train its artificial intelligence model Llama. The publishers, along with author Scott Turow, allege in a proposed class-action complaint that Meta pirated millions of their works and used them without permission to train its large language models to respond to human cues.
Maria Palant, president of the Association of American Publishers, said in a statement: "Meta's mass infringement is not public progress, and artificial intelligence will never be properly implemented if technology companies prioritize pirate sites over academic research and imagination."
The publisher accused Meta of stealing everything from textbooks and scientific papers to novels to train its AI, including works by N.K. Jemisin’s Season 5 and Peter Brown’s Wild Bots. They are asking the court for permission to represent a broader group of copyright owners and seek unspecified financial damages.
The case opens a new front in the ongoing copyright battle between creators and technology companies over AI training, with dozens of writers, news organizations, visual artists and other plaintiffs suing companies including Meta, OpenAI and Anthropic over copyright infringement.
The core dispute in all pending cases is likely to be whether AI systems use copyrighted materials to create new content that is transformative (transformative means significantly innovative or creative), thus constituting fair use. Last year, the first two judges to hear the case issued dramatically different rulings.
Anthropic, backed by Amazon and Google, was the first major AI company to settle one of the cases, agreeing last year to pay $1.5 billion to a group of authors to resolve a class-action lawsuit. If it loses, the company could face billions of dollars in additional damages for alleged infringement.