Google is facing another class action lawsuit over how it trained its Gemini AI models. A coalition of major publishers and authors filed the complaint on Tuesday, alleging the company used copyrighted works without permission and deliberately removed copyright information to hide the infringement.
The plaintiffs include Hachette, Cengage, Elsevier, author Scott Turow, and the nonprofit S.C.R.I.B.E. The group claims Google copied their books and articles to train Gemini, then stripped or altered copyright metadata to conceal the source. The lawsuit describes this as a deliberate attempt to train AI on stolen materials.
A growing wave of copyright complaints
<
p>This case joins a long list of similar lawsuits filed against AI companies such as Meta, OpenAI, and Anthropic. Many are still moving through the courts. Two early rulings in California sided with the technology companies, finding that training AI on copyrighted works qualifies as fair use under a copyright law that predates the internet. Those decisions are not binding nationwide, but they show how courts may view the defense.
However, the landscape is not entirely favorable to AI firms. Anthropic was ordered to pay $1.5 billion for using pirated works in its training data, the largest copyright payout in U.S. history. Around 500,000 writers were eligible for at least $3,000 each, though many opted out to preserve their right to sue over AI training directly.
The Google case was filed in the U.S. District Court for the Southern District of New York. That gives a different judge the chance to weigh in on the fair use question, and the outcome could differ from the California rulings.
Google’s unique relationship with publishers
The lawsuit highlights a more complex history between Google and publishers. For years, publishers provided Google with copyrighted works for the Google Books program, which lets users search for books and see short snippets plus bibliographic data. The program does not allow full book access. The plaintiffs argue that Google took copies from that program and from the Google Play store and used them to train Gemini without explicit permission.
“Google illegally copied works from all these scope-limited programs for AI training, knowing it lacked authorization to do so,” the lawsuit states.
The filing also cites an internal Google document that reportedly warns that using copyrighted books for AI training could be “highly problematic for Google” and could lead to fines in the range of tens to hundreds of billions of dollars.
If the courts side with the publishers, the case could reshape how AI companies obtain training data. Google may have to prove it did not use copyrighted material beyond the narrow permissions granted for search. For publishers, the fight is about protecting the value of their work in an era where AI can replicate it without compensation. For context on the broader trends in humanoid robotics, see our coverage of the humanoid robot race.
