Major publishers sue Google over alleged use of copyrighted books in Gemini training

Google is facing another significant AI copyright lawsuit, this time from a group of major publishers and authors who allege the company used copyrighted books without authorization to train its Gemini models.

The plaintiffs include Hachette, Cengage, Elsevier, author Scott Turow, and S.C.R.I.B.E. According to the complaint, Google not only trained AI systems on protected works without permission, but also altered or removed copyright management information in ways meant to obscure the origin of the material.

A new front in the AI copyright battle

The case adds to a widening wave of litigation against AI developers including Google, OpenAI, Meta, and Anthropic. At the center of these disputes is a question that has become one of the most important legal issues in generative AI: whether using copyrighted content to train large models is lawful under existing U.S. copyright doctrine, especially fair use.

Recent court decisions in California have given some support to AI companies by finding that certain training uses can qualify as fair use. But those rulings have not settled the issue across the industry, and they leave room for other courts to interpret the facts differently. The Google case, filed in the U.S. District Court for the Southern District of New York, could therefore become another closely watched test of how judges assess AI training practices.

Why publishers may frame this case differently

What makes this lawsuit notable is the profile of the plaintiffs and the publishing sector's long-standing licensing relationships with technology platforms. Publishers are not just arguing that Google copied works at scale; they are also likely to emphasize the existence of established markets for licensed digital content, educational materials, and searchable text archives. That could strengthen an argument that unauthorized AI training harms a real and existing market, a key issue in fair use analysis.

The allegation that Google removed or modified copyright information could also raise the stakes beyond a standard training dispute. Claims tied to copyright management information often suggest deliberate concealment rather than mere incidental copying, which may give plaintiffs additional leverage if the court allows those arguments to proceed.

Why this matters for the AI industry

The outcome of these cases will shape more than liability for one company. They could influence how foundation model developers source training data, whether licensing becomes a standard cost of building frontier AI systems, and how rights holders negotiate with tech platforms going forward.

For Google, the lawsuit creates fresh legal pressure around Gemini at a time when the company is competing aggressively in generative AI. For publishers and authors, it is part of a broader push to ensure that AI firms cannot treat copyrighted books and journals as free raw material for model development.

Even if this case takes years to resolve, it underscores a larger shift: AI copyright litigation is moving from isolated creator complaints to coordinated actions by major institutional rights holders. That makes the dispute not just another lawsuit, but a sign that the commercial rules of AI training are still being contested in court.