Meta and its chief executive Mark Zuckerberg are sued by a coalition of five major publishers—Hachette, Macmillan, McGraw Hill, Elsevier and Cengage—along with author Scott Turow. They allege the company illegally scraped millions of copyrighted books and journal articles from sites hosting pirated material to train its Llama AI models for future applications across sectors.

The lawsuit claims Meta accessed the content without permission, then reproduced and distributed it, stripping attribution data to conceal training sources. Plaintiffs argue that Llama now produces imitation versions of their works, flooding Amazon and displacing human‑authored books. They seek unspecified damages and broader representation for copyright owners in the global publishing market today and and.

Meta counters that it pursued licensing talks before abandoning them at Zuckerberg’s instruction and maintains the training falls under fair use. The company points to a June ruling that AI use of copyrighted text can be transformative and notes courts have upheld similar defenses for Microsoft, OpenAI and Anthropic, which paid $1.5bn to settle a claim.

If the court sides with the publishers, Meta could face substantial liability and a precedent that forces AI firms to negotiate licenses or face costly litigation. The case underscores the growing clash between tech giants and intellectual‑property holders, signalling that the industry must address how copyrighted material fuels AI training without compromising creators’ rights for future innovation today and growth.