Google faces collective copyright lawsuit over Gemini AI training

Editors, publishers and author Scott Turow have filed a collective lawsuit in the U.S. District Court for the Southern District of New York, accusing Google of copying millions of copyrighted books and scientific articles to train its Gemini artificial‑intelligence model. The complaint names major publishers Hachette Book Group, Cengage Learning and Elsevier, and alleges that Google used the material beyond the limited “snippet” permissions granted under Google Books, Google Play and Google Scholar agreements.

Internal Google audits cited in the filing indicate that company staff were aware the practice violated copyright law, and a confidential memo estimated potential penalties between $10 billion and $100 billion. The case follows earlier litigation against Google and other tech firms over generative‑AI training data, and could force the courts to define whether large‑scale use of copyrighted works for AI training constitutes piracy.

The lawsuit also references a prior 2023 suit by illustrators and writers, and notes that competitor Anthropic has already paid $1.5 billion to settle similar claims, highlighting the broader industry pressure on AI developers to obtain licensing agreements for source content.