A coalition of major publishing houses — including Cengage, Elsevier, McGraw-Hill, Macmillan, and Hachette — has filed a federal lawsuit against Meta in Manhattan. They accuse the tech giant of unlawfully harvesting millions of copyrighted works to train its Llama AI models. Author Scott Turow is also among the plaintiffs in the proposed class action.
Did Meta Pirate Content to Train its AI Models?
The complaint alleges that Meta scraped a sweeping range of material without permission. The lawsuit states that the company has used everything from academic textbooks and scientific journals to popular novels. Allegedly, they then use pirated content as fuel for their large language models. The publishers are seeking both class certification — to represent a broader pool of rights holders — and an unspecified sum in monetary damages.
Copyrighted Material and Fair Use
Meta has pushed back firmly, with a spokesperson asserting that training AI on copyrighted material can qualify as fair use under existing law. The company pledges to contest the lawsuit aggressively.
Creatives vs. AI Developers in AI Training Battle
The case is the latest skirmish in a rapidly expanding legal war between the creative community and AI developers. Dozens of authors, journalists, visual artists, and media organizations have brought similar infringement claims against companies including Meta, OpenAI, and Anthropic. The central legal question in all these cases is whether AI systems sufficiently transform source material to qualify for fair use protections. This is a question courts have so far answered inconsistently. The first two judges to weigh in issued conflicting rulings in 2025.
Anthropic Settles with Authors for $1.5 Billion
One notable precedent has already been set: Anthropic became the first major AI company to settle such a case. They agreed to pay a group of authors $1.5 billion in 2025 to resolve a class-action that could have carried far steeper financial consequences.
No consensus has yet emerged from the courts. The outcome of the Meta lawsuit could significantly shape how copyright law applies to AI development going forward.