According to the lawsuit, books licensed to Google Books, Play Books, and Scholar were meant for snippets and ebook sales, but Google allegedly used them to train its commercial AI models.
A consortium of book publishers has launched a complaint against Google, saying that the firm illegally exploited millions of copyrighted books to train and improve its Gemini AI models.
The action was brought in federal court in New York by three publishers, Hachette Book Group, Cengage Learning, and Elsevier, as well as bestselling American novelist Scott Turow.
According to the lawsuit, the books were licensed for use on Google Books, Google Play Books, and Google Scholar. Furthermore, the content was intended to be used to show searchable snippets or sell ebooks, but the corporation utilized it to train commercial AI products.
The complaint states, “Desperate to maintain its online dominance, Google abandoned its early motto of ‘Don’t be evil’ and engaged in one of the most prolific infringements of copyrighted materials in history.”
The plaintiffs further claim that the corporation duplicated copyrighted works many times throughout the AI training process without permission or compensation.
The complaint also claims Google deleted copyright management information from copyrighted works “to conceal its training sources and facilitate their unauthorised use.”
According to the petition, Google’s AI models may create content that competes with original works, such as summaries, substitute textbook chapters, alternative novel versions, and prose that resembles certain authors’ styles.
The complaint also argues that Google was aware of the legal concerns connected with employing copyrighted literature for AI training. It claims that business conversations acknowledged the risk of “($10Bs-$100Bs in potential fines)” for using publisher-supplied materials.
The plaintiffs are demanding statutory damages, a permanent injunction to prevent the alleged infringement, and a court order for Google to erase illegal copies of copyrighted works used to train its AI systems.
The case adds to a rising number of copyright cases against AI companies, including Google, OpenAI, Anthropic, and Meta, for using copyrighted content to train generative models.
