Chicken Soup for the Soul sues Apple, Google, and five others over AI training data theft

Publisher Sues Tech Companies Over AI Training Data

Chicken Soup for the Soul has launched an AI lawsuit against seven major technology companies, accusing them of copyright infringement for allegedly using pirated copies of its books to train their artificial intelligence systems. The lawsuit, filed in California federal court, targets Apple, Google, Nvidia, Meta, OpenAI, Anthropic, Perplexity AI, and Elon Musk's xAI 1

1

. The complaint represents one of the most sweeping legal challenges yet in the ongoing battle over AI training data, taking aim at multiple tech juggernauts simultaneously.

The publisher, whose inspirational book series has sold more than 500 million copies worldwide, alleges that these companies illegally used pirated copies of books downloaded from shadow libraries including The Pile, LibGen, Z-Library, and Anna's Archive 1

1

. According to the lawsuit, defendants "helped themselves to the copyrighted works of thousands of authors" to train their large language models and chatbots 2

2

.

Why Chicken Soup Content Matters for LLMs

The lawsuit argues that Chicken Soup for the Soul's content is uniquely valuable for training AI to replicate human communication. The publisher's first-person narratives in "natural, conversational language that conveys emotion, moral reflection, and coherent storytelling in concise form" are particularly suited to teach AI systems to "replicate authentic human voice, narrative pacing, emotional tone, and story structure" 1

1

.

"Instead of paying for that value or licensing access to these works, Defendants pilfered illegal copies and used those copies to build systems now worth many hundreds of billions of dollars," the complaint states 1

1

. Kyle Roche, partner at Freedman Normand Friedland representing the publisher, emphasized that "companies cannot build billion-dollar technologies on stolen creative expression" 1

1

.

Apple's Defense: Research vs. Production Use

The lawsuit specifically alleges that Apple Foundation Models relied upon The Pile dataset to train their artificial intelligence systems 2

2

. However, Apple has previously stated that The Pile dataset was only used for research purposes and not actually deployed in any models powering Apple Intelligence or machine learning features 2

2

.

This distinction between research and production use could prove critical as the case unfolds. Whether courts will view this as a meaningful difference or merely a technical distinction remains to be seen, but it highlights the complex legal questions surrounding how copyright owners can protect their intellectual property in the AI era 2

2

.

Broader Implications for AI Development

This lawsuit joins dozens of high-stakes cases brought by authors, news outlets, and other copyright owners against tech companies for using their work to train their large language models behind chatbots 1

1

. The same law firm, Freedman Normand Friedland, has brought similar ongoing litigation against Big Tech companies on behalf of writer John Carreyrou and other authors 1

1

.

The outcome of these cases will likely shape how AI companies source training data moving forward. If courts side with publishers and copyright owners, tech companies may face substantial licensing costs or need to fundamentally restructure their approach to training AI models. For AI developers and companies building on these platforms, the legal uncertainty around AI training data represents a significant risk that could affect product roadmaps and business models. Watch for how defendants respond to these allegations and whether any settlements emerge that establish precedent for licensing agreements between publishers and AI companies.