Class Action Lawsuit Alleges Apple Illegally Uses Copyrighted Works for AI Training

New to ClassAction.org? Read our Newswire Disclaimer

A proposed class action lawsuit claims that Apple has used copyrighted works to train its AI language models without obtaining the proper licensing or consent.

Want to stay in the loop on class action lawsuits that matter to you? Sign up for ClassAction.org’s free weekly newsletter.

The 18-page AI lawsuit claims that Apple used a dataset called Books3, which contains text files for approximately 196,640 copyrighted written works, to train its OpenELM generative AI language model variants: OpenELM-270M, OpenELM-450M, OpenELM-1_1B and OpenELM-3B.

According to the complaint, when Apple launched Apple Intelligence, which includes the OpenELM language models, the company advertised that Apple Intelligence was trained using “publicly available” or “open source” works. However, the suit alleges that “publicly available” and “open source” do not mean that the included works were made free for public use by the author, but rather that they are simply accessible online, regardless of the legality of that access.

Related Reading: Class Action Lawsuit Alleges Apple Misrepresented iPhone 16 AI Capabilities

The complaint cites the curator of the Books3 dataset, Shawn Presser, stating that the contents of the dataset represent the entirety of a “shadow library” called Bibliotik. A shadow library, the filing explains, is a sort of online archive that hosts enormous quantities of unlicensed, copyrighted materials.

The complaint alleges that the sources of the dataset and, therefore, Apple possessed the dataset’s contents illegally, and that the implementation of these pirated works into AI language model training is a cut-and-dry violation of U.S. copyright law.

Additionally, the case says that any use of the Books3 dataset within the timeframe described in the lawsuit required the user to download a copy of the dataset, which the filing takes to mean that Apple still maintains its own copy of the pirated library.

Finally, the lawsuit argues that the illegal use of these works to train AI language models has contributed and will continue to contribute to the dilution of the works’ market potential, as AI-generated works have consistently been used to put out quickly-produced, low-quality replicas, companions, or other works based on the stolen work the language models were trained with.

The Apple AI class action lawsuit seeks to represent anyone whose copyrighted work or works were used in or otherwise retained for the training or use of Apple’s AI language models without their consent in the past three years.

Want to learn how to start a class action lawsuit? We’ve got you covered.