Apple is the latest tech company to be hit with a proposed class action lawsuit over unauthorized training of AI models using published books. On Friday, authors Grady Hendrix and Jennifer Roberson filed a lawsuit in Northern California targeting Apple's "OpenELM" large language models, alleging the company "copied protected works without consent and without credit or compensation." The suit accuses Apple of using the Books3 dataset of pirated books and employing its own proprietary Applebot to scrap the web and, potentially, other online "shadow libraries."

Hendrix is a horror writer based in New York whose latest book is Witchcraft for Wayward Girls (Berkley), and Roberson, based in Arizona, is a fantasy author with multiple series spanning more than 120 books.

The legal action comes amid a wave of similar cases targeting major technology companies over their AI training practices. The news of the new lawsuit came on the same day AI company Anthropic disclosed it had agreed to pay $1.5 billion to settle a similar class action lawsuit with authors and their representatives. Lawyers for the plaintiffs called the Anthropic settlement "the largest publicly reported copyright recovery in history," though the company did not admit liability.

The expanding litigation highlights the publishing industry's struggle to protect intellectual property rights as tech companies increasingly rely on vast datasets of published, and in many cases copyrighted, content to develop AI systems. The cases raise fundamental questions about fair use, compensation and consent in the digital age, as authors and publishers seek to establish legal precedents for how their works can be used in AI development.