Three publishing companies, Hachette Book Group, Cengage Learning, and Elsevier, along with best-selling author Scott Turow, on Friday filed a proposed class action lawsuit against Google alleging wilful copyright infringement involving millions of textual works used to develop Gemini large language models.

The plaintiffs are seeking to protect intellectual property rights across fiction, nonfiction, children's books, memoirs, poetry, educational materials, and scholarly articles.

The complaint alleges that Google illegally copied books and journal articles obtained for strictly limited use through Google Books and other services, used unauthorised web content, and trained its AI system on the materials without permission. It further alleges that Google removed copyright management information, concealed training sources, and created a service designed to generate content that could substitute for original works. The filing cites internal Google assessments that identified legal and business risks associated with using publisher-provided books for AI training, including potential copyright claims. The lawsuit aims to preserve publishers' and authors' rights to pursue claims beyond the scope of existing litigation involving Google's generative AI copyright practices.