Mark Zuckerberg personally approved the widespread copying of copyrighted books and articles to train Meta's artificial intelligence systems, a new lawsuit alleges. The claim forms part of legal action brought by five major publishers and the author Scott Turow against the tech giant, escalating the ongoing conflict between the literary world and AI developers.
Elsevier, Cengage, Hachette, Macmillan and McGraw Hill accuse Meta of using pirated copies of millions of their copyrighted works to train its large language model, Llama. The plaintiffs say the infringement was carried out on a mass scale and was authorised at the highest level of the company, with Mr Zuckerberg's explicit approval.
Maria Pallante, president of the Association of American Publishers, said in a statement: "Meta's mass-scale infringement isn't public progress, and AI will never be properly realized if tech companies prioritize pirate sites over scholarship and imagination." The lawsuit is the latest in a series of disputes in which authors, news outlets and artists have sued firms including Meta, OpenAI and Anthropic over the use of their work in AI training.
The central legal question in these cases is whether using copyrighted material to train AI systems constitutes fair use. Two federal judges considering similar matters issued conflicting rulings last year. Amazon- and Google-backed Anthropic became the first major AI company to settle one such case, agreeing to pay a group of authors $1.5bn to resolve a class action over alleged piracy.
The publishers allege that Meta's training data included works ranging from textbooks and scientific articles to novels such as The Fifth Season by N.K. Jemisin and The Wild Robot by Peter Brown. They have asked the court to allow them to represent a larger class of copyright owners and to award unspecified monetary damages. A Meta spokesperson responded: "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."
