Apple Inc. is facing a new class-action lawsuit from a group of authors who allege that the company used their copyrighted works without permission for AI training. Filed on October 22 in the U.S. District Court for the Northern District of California, the case claims that Apple’s language models, including OpenELM and Foundation, relied on vast amounts of illegally obtained text from online repositories. These sources, the complaint says, are notorious for hosting pirated books and creative materials.
Author Tasha Alexander leads the lawsuit, representing herself and other registered copyright owners. The filing accuses Apple of pulling content from “shadow libraries” like Books3, LibGen, and Z-Library, which circulate unauthorized copies of literary works. Two senior Apple executives involved in AI development and software engineering have also been listed as defendants.
The complaint contends that Apple used this unlicensed content to develop its generative AI systems now embedded in iPhones, iPads, and Macs. The plaintiffs assert that Apple neither notified nor compensated the writers whose works allegedly fueled its AI training, calling the act a deliberate breach of the Copyright Act of 1976. They argue that Apple engaged in mass copying for commercial purposes, turning creative property into data for corporate gain.
Under U.S. copyright law, only rights holders can authorize reproduction or distribution of their work. Even when done for AI training, the use of copyrighted material requires explicit permission unless deemed fair use. The authors maintain that Apple’s copying cannot qualify as fair use because it was extensive, profit-driven, and harmed their ability to sell their books.
The lawsuit surfaces amid a wave of similar cases testing how copyright law applies to artificial intelligence. U.S. courts are still grappling with whether the use of creative works for AI training violates existing statutes or falls under fair use. While decisions in other ongoing cases may offer direction, the legal landscape remains uncertain.
According to the complaint, Apple benefited financially from these works but did not compensate the affected authors, despite paying licensing fees to companies such as Shutterstock for image-based training data. This, the plaintiffs argue, proves that Apple recognized the market value of creative content yet selectively avoided paying certain creators.
The proposed class seeks to include all authors and rights holders whose published works were allegedly copied or referenced during Apple’s model development. The plaintiffs are pursuing statutory damages, restitution, and an injunction to bar future unlicensed use of their content in Apple’s AI systems.
