Apple is facing a class action lawsuit filed by authors Grady Hendrix and Jennifer Roberson, accusing the tech giant of using pirated books to train its artificial intelligence systems. The lawsuit, lodged in the U.S. District Court for the Northern District of California, claims that Apple incorporated a dataset named “Books3,” which allegedly contains thousands of pirated books sourced from the shadow library Bibliotik.
The Main Allegations
The plaintiffs assert that the Books3 dataset was part of the RedPajama collection, which Apple utilized in training its OpenELM models in 2024. These models were integral to powering Apple Intelligence, and the lawsuit argues that this could mean Apple’s Foundation Language Models were also trained on pirated materials without permission or compensation to the authors. The plaintiffs demand a jury trial and are seeking financial damages, attorney’s fees, restitution, and even the destruction of any AI models developed using this disputed data.
Industry Context and Legal Precedents
This case brings to mind a recent settlement where Anthropic paid $1.5 billion to authors over AI book piracy. Although Apple is not accused of directly scraping books, the lawsuit raises concerns about the questionable origins of some datasets the company may have used for its AI training.
Apple’s Stance on Ethical AI
Apple has long claimed that it adheres to ethical practices in AI development. The company has previously invested millions in securing rights to content from publishers and image licenses from platforms like Shutterstock. In July 2025, Apple reiterated its commitment to respecting copyright by following robots.txt restrictions on websites, unlike some competitors who have been criticized for circumventing such safeguards.
What’s at Stake
As the case progresses, its outcome could have significant implications for how tech companies gather data for AI training. Apple’s strong public assertions about ethical AI practices will now face scrutiny in the courtroom, and the result may influence industry standards moving forward.