Five major publishers sue Meta over copyright and AI training
the Guardian ● Covered by 9 sources
Five publishers sued Meta, saying it used their books and journals to train Llama. The fight is now over whether AI training on copyrighted work is fair use or plain piracy.
Based on reporting by the Guardian — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
Five major publishers have taken Meta to federal court in Manhattan, accusing the company of using their books and journal articles to train its AI models without permission. The plaintiffs are Elsevier, Cengage, Hachette, Macmillan and McGraw Hill, along with author Scott Turow. They say Meta copied millions of works and fed them into its Llama large language models so the system could answer human prompts.
The complaint, filed as a proposed class action, goes well beyond a single title or two. The publishers say the material Meta used spans textbooks, scientific articles and novels, including The Fifth Season by N.K. Jemisin and The Wild Robot by Peter Brown. They are asking the court to let them represent a broader class of copyright owners and to award monetary damages, though they have not specified an amount.
Meta says it did nothing wrong. A spokesperson said AI is driving productivity and creativity and argued that courts have already found training on copyrighted material can count as fair use. The company also said it will fight the case aggressively. That puts the dispute squarely in the middle of a much larger legal battle over how AI systems are built.
And this is not happening in a vacuum. Dozens of authors, news outlets and visual artists have already sued companies including Meta, OpenAI and Anthropic over similar claims. The central question in those cases is whether using copyrighted material to make AI output is a lawful transformation or just copying with a new label. Last year, the first two judges to weigh in reached different conclusions. Anthropic later became the first major AI company to settle one of these fights, agreeing to pay authors $1.5bn.
My take — AI-written commentary, not fact-checked reporting
Big AI companies have spent years acting like training data is free because computers touched it. Courts are now being asked to decide whether that theory is law or just Silicon Valley with a straight face. The copyright crowd is finally making the industry pay for treating books and articles like spare parts.
Read more about this at: the Guardian