Five major publishing houses — Hachette, Macmillan, McGraw Hill, Elsevier, and Cengage — along with bestselling author Scott Turow have launched a class-action lawsuit accusing Meta and its CEO Mark Zuckerberg of copyright violations.
Filed Tuesday in federal court in New York, the complaint alleges that Meta unlawfully used millions of copyrighted books and academic materials to train its AI system, Llama. It further claims the company stripped away copyright notices and other ownership identifiers from those works.
According to the lawsuit, Meta’s developers allegedly sourced pirated content from sites such as Anna’s Archive, which aggregates material from shadow libraries like LibGen and Sci-Hub. The filing goes as far as to claim that Zuckerberg personally approved and encouraged these actions.
Meta has not issued a public response to the allegations.
The plaintiffs argue that AI systems like Llama pose a direct threat to the publishing industry by enabling the rapid creation of imitation books and detailed summaries that could replace original works. They point to a surge of AI-generated titles appearing on Amazon, suggesting these outputs are crowding out human-written content.
The complaint references several authors whose works were allegedly used in training, including V.E. Schwab, N.K. Jemisin, Lemony Snicket, and Turow himself.
As part of its evidence, the lawsuit cites responses generated by Llama. In one example, the AI reportedly produced a travel guide mimicking the style of Becky Lomax. When asked how it achieved such accuracy, the system allegedly indicated it had been trained on a wide range of texts, including her published work.
In another instance, Llama was said to have generated a detailed summary of Turow’s novel Presumed Innocent, even acknowledging it had been trained on a digital version of the book.
Turow criticized Meta’s actions, describing them as reckless and harmful. He expressed frustration that one of the world’s largest corporations would knowingly rely on pirated versions of his work and others’ to build a system capable of producing competing content in their style.
The lawsuit argues that such AI-generated outputs could erode the value of original writing by replicating key elements like plot structure, characters, and themes closely enough to substitute for the real thing.
This case is part of a broader wave of legal challenges from authors and publishers seeking to curb how tech companies use copyrighted material in AI development. Similar lawsuits have been filed against firms including OpenAI, Anthropic, Google, and xAI. Notably, Anthropic previously agreed to a $1.5 billion settlement over comparable claims.
Meta has faced related legal scrutiny before. In 2025, a court ruled in its favor, stating that earlier plaintiffs had failed to prove that its AI tools would significantly harm the market for books.
The new lawsuit brings together a broad coalition of trade, educational, and academic publishers alongside a prominent author. They are seeking a court order compelling Meta to delete any unlawfully obtained materials used in training its AI systems, halt further violations, and provide additional remedies as deemed appropriate.
Maria A. Pallante, CEO of the Association of American Publishers, emphasized the need for a more balanced AI ecosystem — one that ensures transparency, fairness, and safeguards for creators. She warned that the negative impact on authors and publishers is already becoming clear.
