Google Faces New AI Training Lawsuit From Hachette, Cengage, and Elsevier
Major publishers and author Scott Turow accuse Google of training Gemini on millions of copyrighted books without permission, adding another front to the AI industry's copyright battles.
Google is facing fresh legal pressure over how it built Gemini.
Hachette Book Group, Cengage Learning, Elsevier, and bestselling author Scott Turow filed a proposed class action against the company Friday in the U.S. District Court for the Southern District of New York.
The filing accused Google of training its AI model on millions of copyrighted books and journal articles without permission. According to The Guardian, the same group of publishers argues that this practice weakens the business model that supports the book and academic publishing industry.
From Google Books to Gemini’s Training Data
The complaint arrives just a day after Google faces a Swiss antitrust probe, focusing on how Google allegedly obtained the material.
According to TechCrunch, publishers and authors have long provided copyrighted works for Google Books, which only displayed short snippets and bibliographic details instead of full text.
The plaintiffs claim Google instead trained Gemini on copies of those same books, plus titles uploaded to the Google Play Store, without ever securing permission for AI training.
The lawsuit also claims Google intentionally removed or changed copyright information embedded in the works to hide that Gemini had been trained on what the plaintiffs describe as stolen material.
It also cites an internal Google document that reportedly warned this type of training would be “highly problematic” for the company and could expose it to tens or even hundreds of billions of dollars in fines.
Part of a Wider Wave of AI Copyright Fights
This is not an isolated case. This lawsuit is part of a growing wave of legal battles from publishers, authors, and other rights holders against Google, OpenAI, and Anthropic over AI training practices.
The same four plaintiffs also sued Meta, which is aiming to rival top AI firms, earlier this year over similar allegations.
The Guardian says the Google lawsuit reflects a broader effort by publishers, who now see AI training as a serious threat to their industry rather than a one-time dispute.
Anthropic has already faced major financial consequences, with a $1.5 billion copyright payout over pirated training material for its Claude model.
Around half a million authors became eligible for at least $3,000 each, though many chose to reject the settlement and continue pursuing their own legal claims.
Why the New York Venue Could Matter
Two earlier California rulings favored AI companies, finding that training AI models on copyrighted content qualifies as fair use under existing copyright law.
California has also backed AI through a landmark deal to deploy Anthropic’s Claude across state and local public agencies.
However, TechCrunch notes those rulings do not apply to other courts. This lawsuit was filed in the Southern District of New York, where a different judge can reach a different conclusion.
The Guardian reports that publishers chose New York as part of a broader strategy to challenge fair use outside California, where AI companies have seen their strongest legal wins.
With Gemini now at the center of Google’s AI business and New York courts yet to rule on these legal questions, the outcome could influence how future AI training lawsuits are decided.
Source: Book publishers sue Google for copyright infringement over Gemini AI training
