Quick Answer
- USA Today Co. (formerly Gannett) sued OpenAI on October 8, 2026 in federal court in Manhattan (Southern District of New York).
- The case covers content from 19 publications, including USA Today, The Tennessean, Detroit Free Press and The Arizona Republic.
- It asks for more than $250 million in damages, a jury trial and an order to stop the alleged copying.
- The claims: copying hundreds of thousands of articles to train ChatGPT, copying through ChatGPT’s summaries, and removing copyright information from articles.
- OpenAI had not responded in the reports we checked. These are allegations, not court findings.
What is the lawsuit about?
USA Today Co. says OpenAI used its news articles to train ChatGPT without permission. The case is called USA Today Co., Inc. v. OpenAI Foundation. Several local papers owned by the company are co-plaintiffs, and they say they own the copyrights in the articles at issue.
The company names seven OpenAI entities as defendants, including OpenAI Foundation, OpenAI Group PBC and OpenAI OpCo. It also asks that the case be treated as related to the large group of OpenAI copyright cases already in the same court.
The three main claims
According to a summary of the complaint, the publishers make three types of claim:
- Training copyright claim. OpenAI allegedly copied hundreds of thousands of articles to train its models, including content behind paywalls.
- Output copyright claim. The complaint says ChatGPT’s summaries (it names GPT-5.6) follow the structure of the original articles and can replace them.
- Copyright information removal. The publishers say OpenAI used programs that strip details such as author names and notices from articles.
What numbers does the complaint cite?
These figures come from one summary of the complaint, so check them against the court record:
- More than 160,000 entries from the publishers’ sites in a data set called WebText, including 83,266 from usatoday.com and 12,994 from freep.com.
- More than 122 million tokens from the publishers in a data set called C4.
- The complaint says the models involved range from GPT-1 to GPT-6.1 and GPT-OSS.
- It quotes OpenAI’s own numbers: over 900 million weekly ChatGPT users and more than 50 million paying subscribers as of March 2026.
What does the publisher want?
- Damages “in excess of $250 million.”
- Under US law, statutory damages can reach up to $150,000 for each willful infringement and up to $25,000 for each removal of copyright information. The complaint relies on those limits.
- A jury trial and a court order blocking the alleged copying.
Why is this case different?
Many news companies already sue AI firms. The New York Times, The Intercept, Ziff Davis, CBC/Radio-Canada, Encyclopaedia Britannica and Merriam-Webster have all brought cases against OpenAI, per one report. What stands out here:
- Local news at scale. The plaintiff owns a very large group of local newspapers, so the case speaks for many small publishers.
- Output claims, not only training. The complaint argues that ChatGPT’s answers can stand in for reading the article, which goes beyond the usual “you trained on our work” argument.
- Internal quotes. The complaint reportedly cites statements from OpenAI staff, such as a research leader saying “we train our networks to memorize the training data.” A summary of the complaint also says a ChatGPT leader said there was “no good reason to click” a source link. Treat both as the plaintiff’s claims.
What does OpenAI say?
No statement from OpenAI appeared in the reports we checked. In other copyright cases, OpenAI and other tech companies have argued that training can count as fair use. USA Today Co. argues the opposite for its news content.
What happens next?
OpenAI has to respond in court. The request to link it with the existing OpenAI copyright cases in Manhattan could decide which judge handles it. Cases like this can take years, so no quick result should be expected. For other OpenAI news, see our report on OpenAI DevDay and GPT-6.1 Sol, the new ChatGPT Intelligent UI and why OpenAI paused training.
FAQs
Who is suing OpenAI?
USA Today Co., formerly Gannett, and newspapers it owns.
When was the lawsuit filed?
October 8, 2026, in the Southern District of New York.
How much money do they want?
More than $250 million, plus a jury trial and a court order to stop the alleged copying.
Which newspapers are involved?
Content from 19 publications, including USA Today, The Tennessean, Detroit Free Press, The Arizona Republic, IndyStar, The Columbus Dispatch and The Oklahoman.
Has OpenAI responded?
Not in the reports we checked.
Does this mean OpenAI broke the law?
No. A court has not ruled. These are allegations.
