Major Publishers Sue Google Over Alleged Gemini AI Training on Copyrighted Works

A coalition of prominent publishers and authors has filed a class action lawsuit against Google, accusing the technology company of copying their copyrighted works to train its artificial intelligence platform, Gemini. The plaintiffs include major publishing houses Hachette, Cengage, and Elsevier, along with bestselling author Scott Turow and the organization S.C.R.I.B.E.

The complaint, filed in the U.S. District Court for the Southern District of New York, alleges that Google went beyond unauthorized use of the materials. According to the lawsuit, the company intentionally removed or altered copyright information attached to the works in an effort to hide that its Gemini models were trained on what the plaintiffs describe as stolen content.

A Complex History Between Google and Publishers

What sets this case apart from other AI-related copyright disputes is the longstanding relationship between Google and the publishing industry. For years, publishers and authors have supplied Google with copyrighted books for a specific and limited purpose: making those books searchable through Google Books. That program allows users to find books through search queries and view short snippets of text along with bibliographic details, but it does not permit access to entire works.

The plaintiffs now claim that Google took copies of the books provided through this scope-limited program and used them to train Gemini, despite never receiving permission to do so. The lawsuit further alleges that Google also used books uploaded to the Google Play store for AI training purposes without authorization.

According to the complaint, Google copied works from these programs for AI training while fully aware that it lacked the right to do so. The plaintiffs cite an internal Google document that reportedly acknowledged the risks of using copyrighted books for AI training, describing the practice as potentially highly problematic for the company and warning that it could result in tens to hundreds of billions of dollars in potential fines.

A Broader Wave of Litigation Against AI Companies

The lawsuit against Google is part of a much larger pattern of legal challenges brought by publishers, authors, and other copyright holders against companies developing artificial intelligence systems. Google, Meta, OpenAI, and Anthropic have all faced similar complaints from creators who argue that their copyrighted works were used without consent to build AI models.

While many of these cases remain unresolved, two early court decisions in California have sided with the AI companies. Judges in those rulings determined that the use of copyrighted works for AI training qualifies as fair use under existing U.S. copyright law, which has not been updated since before the internet became a widespread tool. These decisions have raised concerns among copyright holders about how courts may view the fair use defense going forward.

However, the legal landscape remains complex. In a separate case, Anthropic was fined $1.5 billion for pirating the works it used to train its AI systems, representing the largest payout in the history of U.S. copyright law. Under that settlement, approximately half a million writers became eligible to receive payments of at least $3,000 each. Notably, many authors chose to opt out of the settlement entirely, preserving their right to pursue additional legal action over AI training practices.

What This Case Could Mean for the Future

The California rulings that favored AI companies may not dictate the outcome of every future case. Legal observers note that the issues involved are too nuanced for any single decision to establish an unassailable precedent. By filing in the Southern District of New York, the plaintiffs have ensured that a different judge in a different jurisdiction will have the opportunity to evaluate the arguments.

The existing relationship between the publishers and Google adds another layer of complexity. The plaintiffs are not simply accusing Google of scraping publicly available content; they are asserting that Google repurposed materials shared for one specific use case, the Google Books search program, for an entirely different and unauthorized purpose. This distinction could prove significant as the court weighs whether Google's actions fall under fair use or constitute copyright infringement.

The internal document cited in the lawsuit, if authenticated, could also play a pivotal role. A company's own assessment that its activities might be highly problematic and could lead to enormous financial penalties may undermine arguments that the practices were conducted in good faith.

Google did not immediately respond to a request for comment on the lawsuit.

As courts across the United States continue to grapple with the intersection of artificial intelligence and intellectual property law, cases like this one could help shape the rules that govern how AI companies build their models. The outcome may have far-reaching implications for publishers, authors, and the technology industry alike. If you found this article informative, please share it with your network and join the conversation about the future of AI and copyright.

Source: TechCrunch