Five major publishing firms and a prominent author filed suit against Meta Platforms on Tuesday in Manhattan federal court, accusing the social media giant of using their copyrighted content without authorization to train its Llama artificial intelligence model.
The proposed class-action complaint lists Elsevier, Cengage, Hachette, Macmillan and McGraw Hill, along with author Scott Turow, as plaintiffs. It alleges Meta copied millions of works belonging to those publishers and the author to build and tune its large language models.
According to the complaint, the material used by Meta for AI training ranged from textbooks and scientific articles to fiction for children and adults. The filing specifically cites titles such as "The Fifth Season" by N.K. Jemisin and "The Wild Robot" by Peter Brown as examples of works the plaintiffs say were incorporated into training datasets.
Maria Pallante, president of the Association of American Publishers, said in a statement that Meta’s actions do not constitute public progress and that AI development should not prioritize unauthorized sources over legitimate scholarship and creative works.
The plaintiffs are asking the court to approve their request to act on behalf of a broader class of copyright owners affected by the alleged copying. They are also seeking monetary damages, though the complaint did not specify an exact dollar amount for those damages.
The filing follows last year’s resolution of a related class-action matter involving Anthropic, a company backed by Amazon and Google. In that case, Anthropic agreed to pay a group of authors $1.5 billion to settle claims related to training on copyrighted material.
The new complaint targets Meta’s Llama models and frames the dispute as a claim that unauthorized use of copyrighted works was central to the company’s AI development. The litigation seeks to establish remedies for rights holders and to secure compensation for the alleged unauthorized use of their works.
Context and implications
The complaint centers on alleged unauthorized copying of copyrighted texts for machine learning training. Plaintiffs are pursuing class certification and monetary relief; the amount of damages sought remains unspecified. The Anthropic settlement cited in the complaint provides one recent example of litigation over similar practices and a large monetary resolution.