A group of publishers and authors have sued Google in the US District Court for the Southern District of New York, alleging the company used copyrighted books without permission to train its Gemini AI models and altered or removed copyright information to hide the source material.

The plaintiffs include Hachette, Cengage, Elsevier, author Scott Turow, and S.C.R.I.B.E.

The complaint said Google trained Gemini on books supplied through Google Books and on titles uploaded to Google Play, even though those programs did not authorise the works for AI training.

Google Books lets users search books and view short snippets rather than full texts.

The lawsuit is part of a broader wave of copyright cases against Google, Meta, OpenAI, and Anthropic over the use of copyrighted material to develop AI models.

The case also comes as US courts begin weighing fair-use defenses in AI copyright disputes. Two early rulings in California found that AI training on copyrighted works can qualify as fair use under some circumstances.

🔗 Source: TechCrunch