Three publishers, novelist Scott Turow, and his company SCRIBE have filed a proposed class action lawsuit accusing Google of copying millions of books and journal articles to train Gemini. This includes works provided through Google Books, Play Books, and Scholar.
On July 10, Hachette Book Group, Cengage Learning, Elsevier, Turow, and SCRIBE joined together in a lawsuit filed in the US District Court for the Southern District of New York, and the Association of American Publishers announced that day. They argued that the books and articles provided by these services were designed for specific purposes, and using them to train a commercial AI model was not one of them. The lawsuit also alleges that Google copied the works it found in web scrapes, including pirate sites and paid libraries. Google did not comment on the complaint as of publication, and no court has ruled on any of the claims. The main question is whether the single-use permit includes training the model on that data.
What the Complaint Means
The complaint brings four charges. The three allege unauthorized reproduction under the Copyright Act, including Google Books and other Google services, web downloading, and copying during training. Fourth, you claim that Google has removed copyright management information in violation of the DMCA. The plaintiffs are asking for damages, an injunction, a detailed account of the works Gemini used for training, and court orders to remove any unauthorized copies. The filing cites what it describes as internal Google documents, one of which called Google Play Books’ use of AI as “very problematic for Google,” with potential fines ranging from “$10Bs-$100Bs.” It reveals another line from Gemini’s lead engineer, who allegedly told his colleagues, “we don’t do deals with data that we already have or already have.” None of these documents are public, and the quotes are from the plaintiffs’ filing.
Where the Crawler Controls Stand
Google-Extended is a robots.txt token that includes the content that Google crawls on your site. It restricts that content to be used for future Gemini training and certain basic uses. Neither of the two ways of getting help discussed here includes that sign. Books are provided directly to Google by agreement, so the robots.txt file does not affect this process. The web-scraping claims refer to copies that, according to the complaint, appeared on Common Crawl after being hosted on pirate sites and registered libraries. Since these copies are hosted on separate domains, the robots.txt file cannot control them.
On June 25, Google published a policy paper arguing that training on public web data is a “transformative, non-disclosure use” under fair use protections. The paper also mentions machine-readable controls, such as Google-Extended, that websites can use to log out. However, the materials examined here are said to have arrived through different channels.
Last month, Next Digital Content sent a cease-and-desist letter to the Common Crawl Foundation, arguing that copyright law doesn’t work as an opt-out.
Why This Matters
The question of consent and the question of fair use are separate concerns. Fair use can apply even if there is no agreement authorizing the use, and an appeal does not solve any problem.
Your browser settings are the least likely factor in this situation. In January, BuzzStream data showed that 79% of top news sites are blocking at least one AI training bottleneck, which is the Extended Addresses channel on Google. The two groups of transcripts analyzed here are said to have arisen from pathways unaffected by those settings.
Looking Forward
In 2025, two Northern California decisions found the use of training consistent with the issues in the records before them. The Anthropic court denied summary judgment for the central library copies, while the Meta judge emphasized that its ruling was specific to those plaintiffs and their record. The publishers said they filed in New York after initially planning to intervene in the ongoing In re Google Generative AI Copyright Litigation in California, and that the new lawsuit preserves claims they believe fall outside that proposed class. The next step is Google’s response, either a response or a motion to dismiss.
