Google is facing a major new copyright lawsuit alleging that it copied millions of copyrighted books and journal articles without permission to train its Gemini family of artificial intelligence models. The lawsuit was filed in the U.S. District Court for the Southern District of New York by major publishers Hachette Book Group, Cengage Learning, Elsevier, along with bestselling author Scott Turow. The plaintiffs seek class-action status on behalf of authors and publishers and claim Google’s actions constitute one of the largest copyright infringements in history.
According to the complaint, Google allegedly repurposed books that had been provided for limited uses through services such as Google Books, Google Play Books, and other Google platforms, using them instead to develop its Gemini AI models. The plaintiffs also allege Google copied content obtained from unauthorized web sources, including pirate websites, without obtaining licenses or compensating rights holders. Google had not publicly responded to the lawsuit at the time of publication.
Publishers Sue Google Over Gemini Training
The lawsuit adds to the growing legal battle over how AI companies obtain training data.
| Key Highlights | Details |
|---|---|
| Company | |
| AI model | Gemini |
| Plaintiffs | Hachette, Cengage, Elsevier, Scott Turow |
| Main allegation | Unauthorized copying of millions of books |
| Court | U.S. District Court, Southern District of New York |
The plaintiffs are seeking damages, injunctive relief, and destruction of allegedly unauthorized copies used for AI training.
What the Lawsuit Alleges
According to the complaint, Google allegedly:
- Copied millions of copyrighted books without permission.
- Used works originally supplied for limited Google services to train Gemini.
- Trained AI models on copyrighted books and journal articles.
- Used material allegedly obtained from pirate sources and web scrapes.
- Built commercial AI products without licensing the underlying works.
The plaintiffs argue that Gemini-generated content can directly compete with original books and educational materials.
Why Publishers Object
The publishers contend that:
- Their works were licensed only for limited purposes such as search previews or ebook distribution.
- AI training represents an entirely different commercial use.
- Authors and publishers should be compensated if their works are used to train AI.
- Gemini can generate summaries, explanations, and other content that may reduce demand for original publications.
Internal Documents Cited
| Allegation | Details |
|---|---|
| Internal warnings | Complaint cites documents warning AI use of publisher books was “highly problematic” |
| Legal risk | Complaint alleges employees discussed potential multibillion-dollar liability |
| Licensing | Plaintiffs argue Google knowingly avoided licensing content it already possessed |
These allegations come from the plaintiffs’ complaint and have not been proven in court.
Part of a Broader AI Copyright Battle
The lawsuit joins a growing wave of legal challenges against AI developers.
Recent disputes involve allegations that AI companies trained models using copyrighted:
- Books.
- News articles.
- Images.
- Music.
- Scientific publications.
Courts in the United States continue to evaluate whether AI training qualifies as fair use under copyright law, with different cases reaching different procedural outcomes depending on their facts.
What Google May Argue
Although Google has not publicly responded to this specific lawsuit, technology companies facing similar cases have generally argued that:
- AI training is transformative.
- Model training does not reproduce original works in the traditional sense.
- Certain uses may qualify as fair use under U.S. copyright law.
Google’s legal position in this case will become clearer once it files its response in court.
Challenges Ahead
The case raises several important legal questions.
These include:
- Whether AI training constitutes fair use.
- Whether previously licensed content can be reused for AI training.
- How copyright law applies to foundation models.
- Whether AI developers must obtain licenses from publishers.
- Potential damages if infringement is established.
The litigation is expected to take months or years before reaching a final resolution.
Outlook
The lawsuit against Google represents one of the most significant copyright challenges yet facing the generative AI industry. By alleging that Gemini was trained using millions of copyrighted books and journal articles without authorization, the publishers are testing whether existing copyright law permits AI companies to repurpose works originally supplied for other services. The outcome could influence how AI developers acquire training data and negotiate licensing agreements with publishers in the future.
Because the case is at an early stage, the allegations remain unproven, and Google will have an opportunity to respond in court. Regardless of the eventual outcome, the lawsuit is likely to become an important precedent in defining the relationship between copyright law and generative AI.
What It Means for the AI Industry
The dispute highlights the growing tension between rapid AI development and intellectual property rights. Publishers and creators are increasingly seeking compensation for the use of copyrighted material, while AI companies argue that model training is a transformative process that can foster innovation.
The eventual court decisions in cases like this could shape future licensing markets, determine how foundation models are trained, and influence the balance between innovation and copyright protection across the AI ecosystem.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
