Publishers sue Google over Gemini training

Hachette Book Group, Cengage Learning, Elsevier, and author Scott Turow filed a putative class action against Google on July 13, in Manhattan federal court (the Southern District of New York), on behalf of a proposed class of authors and publishers. The suit alleges Google trained its Gemini models on millions of copyrighted books and journal articles without permission, drawing on Google Books, Google Play Books, and Google Scholar, plus material the complaint describes as pulled from pirate sites.

The complaint argues that publishers who supplied works to Google Books years ago authorized only searchable snippets, not use as AI training data. It also alleges Google stripped or altered copyright management information embedded in the works to hide where the data came from, a claim that points toward the Digital Millennium Copyright Act alongside the core infringement claims. Plaintiffs cite an internal Google document said to flag training on publisher-supplied books as carrying “$10Bs-$100Bs in potential fines.” They want statutory damages, an injunction, and an order forcing Google to destroy unauthorized copies.

Why a second filing

The move is partly defensive. Hachette and Cengage had tried to join the existing “In re Google Generative AI Copyright Litigation” case in Northern California as publisher class representatives, but Google contested that motion on timeliness grounds. The publishers say they filed separately in New York to protect claims that could otherwise fall outside that case, adding Elsevier and Turow as named plaintiffs.

It is the latest entry in a run of AI copyright fights, following a sanctions motion in the New York Times’ suit against OpenAI and Apple’s trade secret suit against OpenAI, and it comes less than a year after Anthropic agreed to a $1.5 billion settlement, the largest copyright payout in U.S. history, in a similar authors’ case.