A coalition of major publishers and bestselling author Scott Turow has initiated a copyright infringement lawsuit against Google, alleging the tech giant utilized millions of books and scholarly articles without authorization to train its Gemini artificial intelligence models. Filed in the U.S. District Court for the Southern District of New York, the complaint positions the litigation as a direct challenge to how AI companies source data for large language models, arguing that Google’s practices bypass established licensing markets.
The plaintiffs—a group that includes Hachette Book Group, Cengage Learning, and Elsevier—contend that Google’s AI systems were built by scraping copyrighted materials from sources such as Google Books, online libraries, and various websites. According to the court filing, the plaintiffs claim this activity constitutes “one of the most prolific infringements of copyrighted materials in history,” asserting that the company abandoned its former corporate motto of “Don’t be evil” to secure dominance in the generative AI sector.
Allegations of Unauthorized Data Usage
At the center of the dispute is the mechanism by which Google’s Gemini models learn. The lawsuit alleges that the tech company ingested vast quantities of protected intellectual property without providing compensation or obtaining permission from the rightful owners. By incorporating these works into its training datasets, the plaintiffs argue that Google has effectively created a multibillion-dollar business model built upon the uncompensated labor of authors and publishers.
The legal action highlights the potential economic harm to copyright holders, noting that Gemini now functions as a market competitor to the very materials it used for training. By generating textbook-style explanations, detailed summaries, and other written content, the plaintiffs argue that the AI models diminish the demand for original works. The complaint specifically references textbooks from Cengage and novels authored by Turow, such as Presumed Innocent, to demonstrate how the technology allegedly replicates or repurposes copyrighted content.
The Shift in AI Licensing Markets
The lawsuit underscores a broader tension within the publishing industry regarding the rise of generative AI. As noted in the official court docket, the plaintiffs emphasize that there is an existing and growing market for licensing content to AI developers. By utilizing copyrighted books and journals without entering into these agreements, the publishers argue that Google is unfairly bypassing the standard commercial practices that other entities are beginning to adopt.
Elsevier, which publishes academic journals including The Lancet and Cell, represents the scientific publishing sector in this suit.
Google’s Position and Next Steps
Google has not yet provided a formal response to the specific allegations outlined in the complaint. The plaintiffs are seeking both monetary damages for the alleged infringement and a court-ordered injunction that would prevent Google from continuing to use their copyrighted materials for AI training purposes without proper authorization.
This case joins a growing body of litigation involving tech companies and creators over the use of data in AI development. As the legal process moves forward, the court will likely need to address whether the ingestion of copyrighted works for the purpose of training AI models falls under the doctrine of “fair use” or if it constitutes a violation of copyright law. No dates have been set for a trial, and the case remains in the preliminary stages of the litigation process in the Southern District of New York.
Readers interested in following the progression of this case can monitor the Public Access to Court Electronic Records (PACER) system for future filings, motions, and scheduled hearings. We will continue to provide updates as this legal challenge develops.
Related reading