AI and Copyright: Key US Developments from 2025
Artificial intelligence developers continue to face substantial legal exposure in the United States as copyright litigation accelerates. Over the past several years, more than 50 infringement actions have been filed against major AI companies, with around 30 cases still active following consolidations and early settlements. Many of the remaining disputes are proposed class actions brought by groups of authors, but claims have also been filed by visual artists, news organisations, creators of online video content, photographers, programmers, and music rights holders. Most cases are pending in the Northern District of California and the Southern District of New York, with a smaller number progressing in Delaware and Massachusetts.
Although the lawsuits involve different creative works from novels and technical manuals to sound recordings, images, and code most share the same central allegation, that large datasets used to train generative AI systems were built through wholesale copying of copyrighted material without permission. Earlier claims relating to AI output have largely been narrowed by the courts, but allegations of direct infringement during the ingestion and dataset-building stage continue to survive.
Two Significant Fair Use Decisions
Recent rulings in two U.S. federal cases, Bartz v. Anthropic and Kadrey v. Meta, represent the first substantive judicial assessments of fair use in the context of AI training. They reach different conclusions and, taken together, highlight the unsettled state of U.S. copyright law as applied to machine learning.
In Bartz v. Anthropic, the court held on summary judgment that two categories of copying using lawfully obtained books for training and scanning physical copies purchased by the company could qualify as fair use. However, the court found that the company’s downloading of unauthorised digital copies from pirate sites did not fall within the defence. That portion of the case will now proceed to trial, with the defendant potentially facing significant statutory damages given the scale of the copying and the fact the plaintiff class has been certified. While some early commentary framed the decision as a victory for AI developers, the exposure created by the adverse piracy finding and the court’s own acknowledgement of likely appeal issues suggest a far more complicated picture.
Two days later, in Kadrey v. Meta, another judge reached a similar conclusion on training but emphasised that the analysis was limited by gaps in the plaintiffs’ evidence. Importantly, the court undertook a detailed examination of whether AI systems might displace authors’ markets, a factor that U.S. law considers central to the fair use inquiry. The judge signalled that a more robust evidentiary record could have led to a different outcome in future cases. The court also took an unexpected approach to Meta’s use of pirated books, declining to treat the copying as unlawful in the circumstances presented. That aspect of the judgment is likely to attract scrutiny in later appeals and in other courts.
Key Takeaways for Businesses
For organisations developing or deploying AI systems in the U.S., these cases underline that copyright compliance is becoming a decisive commercial risk. Courts are willing to scrutinise the provenance of training data, and the use of pirated or improperly sourced materials is emerging as a major liability issue that may attract class-wide damages.
At the same time, the early fair use rulings show that outcomes will depend heavily on market-harm evidence, transparency around training datasets, and the specific manner in which training materials were obtained. Businesses operating in the AI ecosystem whether as developers, licensors, or downstream users, should monitor these cases closely, audit their dataset acquisition practices, and ensure they can demonstrate lawful sourcing and robust governance over training data.
How Can Gerrish Legal Help?
Gerrish Legal is a dynamic digital law firm. We pride ourselves on giving high-quality and expert legal advice to our valued clients. We specialise in many aspects of digital law such as GDPR, data privacy, digital and technology law, commercial law, and intellectual property.
We give companies the support they need to successfully and confidently run their businesses whilst complying with legal regulations without the burdens of keeping up with ever-changing digital requirements.
We are here to help you, get in contact with us today for more information.