Apple sued over use of copyrighted books to train Apple Intelligence (www.reuters.com)

A new lawsuit alleges that Apple used copyrighted books without permission to train Apple Intelligence, the company’s generative AI offering. Plaintiffs claim their works were ingested into training data and that Apple failed to obtain licenses or provide compensation, seeking remedies that could include damages and limits on how the model is used or distributed. The complaint raises familiar legal questions about whether large-scale ingestion of copyrighted text for model training constitutes infringement or falls under doctrines like fair use. The case matters to the AI/ML community because it targets core practices in building foundation models: dataset curation, provenance, and the legal status of using copyrighted material for training. A ruling against Apple could force more rigorous vetting of training corpora, broader licensing deals, or technical changes to avoid memorization and verbatim reproduction of copyrighted content. It also increases legal and compliance risk for companies releasing consumer-facing generative systems, potentially shaping disclosure norms, data documentation standards, and model fine-tuning workflows. Developers and organizations should watch for precedent-setting outcomes that could alter how training datasets are assembled, how models are deployed, and what safeguards are required to mitigate copyright exposure.