American authors sue Apple over allegations of training AI with pirated books.
Two American authors have filed a lawsuit against Apple in California, accusing the tech giant of using pirated books to train artificial intelligence (AI) models embedded in its devices.
The lawsuit coincides with news that Anthropic, a prominent US AI start-up, has agreed to pay at least $1.5 (~€1.3) billion to copyright holders who accused it of illegally downloading millions of books.
In their court filing, authors Grady Hendrix and Jennifer Roberson claim that Apple built its AI models, part of its “Apple Intelligence” features, by first amassing a massive library of data.
They argue that this data includes copyrighted works copied without their consent, credit, or compensation, and are seeking to transform their case into a class-action lawsuit.
The authors believe their books are among the materials “scraped” by Apple from what they describe as “shadow libraries” of pirated content. Apple has yet to respond to the claims.
The case is part of a growing wave of lawsuits in the US, where writers, musicians, and publishers are accusing tech companies of unauthorised use of their work to train AI systems, which rely on vast amounts of data.
To counter such lawsuits, major players in generative AI often invoke the legal doctrine of “fair use,” which can limit the enforcement of intellectual property rights.
In the Anthropic case, a California judge ruled that training AI with copyrighted material could fall under fair use. However, the court found Anthropic—developer of the AI model “Claude”—guilty of downloading and storing pirated books instead of purchasing them.
In contrast, a separate federal judge in California sided with Meta, Facebook’s parent company, in a similar case in June. The judge pointed out that plaintiffs could have strengthened their arguments by asserting that AI-generated tools, which rely on writers’ works, are creating competition in the literary market.
These cases highlight the complex and evolving legal questions at the intersection of AI development and intellectual property rights.
