Google faces publishers' Gemini copyright suit

Google is facing a new copyright lawsuit over Gemini, with a group of major publishers and author Scott Turow alleging the company used millions of copyrighted books to train its AI without permission or payment.

The plaintiffs — Hachette Book Group, Cengage Learning, Elsevier, Scott Turow, and S.C.R.I.B.E. — filed the complaint on 10 July in the US District Court for the Southern District of New York and are seeking class-action status. The lawsuit calls Google’s conduct “one of the most prolific infringements of copyrighted materials in history.” The Guardian first reported the claims.

The complaint focuses on works Google obtained through limited-use services including Google Books, Google Play Books, and Google Scholar. The publishers argue those programs allowed searchable snippets or ebook sales, not the copying of books and journal articles to train a commercial AI system.

“Google illegally copied works from all these scope-limited programs for AI training, knowing it lacked authorization to do so.”

The suit also alleges Google relied on web scrapes from “known pirate sources” and from behind paywalls. It brings four counts:

Recommended reading

200+ experts urge action on AI’s economic shock

• Three under the Copyright Act
• One under the Digital Millennium Copyright Act, alleging Google removed or altered copyright information to obscure that Gemini had trained on the works

According to the filing, an internal Google document warned that training on copyrighted books could be “highly problematic” and expose the company to “$10Bs-$100Bs in potential fines.”

What the publishers say Gemini can do

The plaintiffs argue Gemini now competes directly with the works it was trained on. They say its outputs can include near-verbatim copies, replacement textbook chapters, and imitative novels. One example in the complaint says Gemini can generate a 100-page murder mystery in about 20 minutes for 39 cents.

The filing names specific works it says Google used, including N.K. Jemisin’s _The Fifth Season_ and Lemony Snicket’s _Who Could That Be at This Hour?_ It also points to Alphabet’s first $100 billion revenue quarter in October 2025, which the plaintiffs link to Google’s AI business.

The group is seeking statutory damages, a permanent injunction, and an order requiring Google to destroy any unauthorized copies used in training. Google did not respond to requests for comment.

The broader copyright fight over AI training data

The case adds to a widening legal battle over how AI models are trained. OpenAI, Anthropic, and Meta have all been sued by authors and publishers over training data, alongside separate cases involving news outlets and film studios. The same publishers also sued Meta earlier this year.

Two rulings in California last year favored AI companies on fair use, though both judges said future cases could still be decided differently. Anthropic separately agreed to pay $1.5bn to authors over pirated copies — the largest copyright payout in US history.

Some rights holders have signed licensing deals with AI companies instead of going to court. But in this case, the plaintiffs say Google already licensed some content for training while still choosing unauthorized sources. A New York judge will now decide whether those claims hold up.