Core Viewpoint - Two authors have filed a class-action lawsuit against Apple, accusing the company of illegally using a pirated dataset, Books3, to train its AI models, including OpenELM and foundational language models [1][4]. Group 1: Lawsuit Details - The plaintiffs, Grady Hendrix and Jennifer Robertson, claim that Apple utilized the Books3 dataset, which contains numerous copyrighted pirated books, for training its open-source models [1]. - The lawsuit includes six main demands: seeking class-action status, economic compensation (including compensatory damages and restitution), a permanent injunction against Apple's continued infringement, destruction of all infringing AI models and training datasets, and coverage of legal costs by Apple [4]. Group 2: Context of the Lawsuit - This lawsuit arises during a critical period of copyright disputes related to AI training, with Anthropic recently settling a similar case for $1.5 billion, while Meta won a lawsuit based on a "fair use" defense [4]. - The core of the dispute revolves around the applicability of the "fair use" principle, with the authors asserting that unauthorized use constitutes infringement [4].