Google sued by major publishers over Gemini AI copyright infringement

A coalition of major publishers – Hachette Book Group, Cengage Learning, Elsevier – and bestselling author Scott Turow, together with the author‑rights group S.C.R.I.B.E., filed a federal lawsuit in the Southern District of New York on July 14 2026. The complaint alleges that Google copied millions of copyrighted books, journal articles and other textual works from Google Books, Google Play Books, Google Scholar and even pirated‑site scrapes, and used them to train its Gemini large‑language‑model AI without permission or compensation. The plaintiffs say the practice violates U.S. copyright law, the DMCA and the companies’ licensing agreements, which limited the use of the works to snippet display or ebook distribution. Google’s internal documents reportedly flagged the risk of “$10‑$100 billion in potential fines” and described the training as “highly problematic.” The suit seeks statutory damages, a permanent injunction to stop further use, and the destruction of any unauthorized copies. It also highlights Gemini’s ability to generate full‑length books or detailed summaries in minutes for a few cents, which the publishers argue could supplant the market for their original works. The case is part of a broader wave of copyright actions against generative‑AI firms, following similar suits against Meta, OpenAI and Anthropic, the latter of which settled for $1.5 billion.