Elsevier, Cengage, Hachette, Macmillan and McGraw Hill filed a lawsuit against her in Manhattan federal court, accusing the tech giant of misusing their books and scientific articles to train the Llama model. The publishers, as well as author Scott Turow, allege in the proposed class action that Meta “pirated” millions of their works and used them without permission to train its large language models to respond to human queries.
“It’s foreordained,” says Meta
“Artificial intelligence is powering transformative innovation, productivity and creativity for individuals and companies, and courts have rightly held that training AI on copyrighted material can fall under fair use,” a Meta spokesperson said, adding: “We will vigorously defend against this lawsuit.”
The publishers allege that Meta illegally used works ranging from school and university textbooks to scholarly articles and novels, including NK Jemisin’s The Fifth Season and Peter Brown’s The Wild Robot, to train its systems. They are asking the court to allow them to represent a broader class of copyright holders, as well as to award unspecified damages.
“Mass breach by Meta” publishers say
“Meta’s massive piracy is not an advance in the public interest, and artificial intelligence will never be fully exploited if tech companies prioritize pirated sources over scientific knowledge and creative imagination,” said Maria Payante, president of the Association of American Publishers.
The lawsuit opens a new front in the ongoing copyright dispute between creators and tech companies over the training of artificial intelligence systems. Dozens of authors, media, visual artists, and other plaintiffs have already sued companies like Meta Platforms, OpenAI, and Anthropic for copyright infringement.
All pending cases are expected to be about whether AI systems make fair use of copyrighted material when they leverage it to create new, transformative content. The first two judges to hear the issue issued conflicting rulings last year.
Anthropic has reached a compromise before
Anthropic, which is backed by Amazon and Google, was the first major AI company to reach a settlement in one of those cases, agreeing last year to pay $1.5 billion to a group of authors to settle a class action that could have cost it billions more in damages for alleged piracy.
