USA TODAY Sues OpenAI for $250 Million Over 19 Newspapers
ChatGPT
USA TODAY Co. and 18 local papers sued OpenAI for more than $250 million, alleging GPT models were trained on their articles without permission and ChatGPT summaries replace reading the originals.
AP
Staff
October 10, 2026 · 3 min read
USA TODAY Co., the publisher formerly known as Gannett, sued OpenAI in federal court in New York on October 8, 2026, seeking more than 250 million dollars. The complaint says OpenAI trained its GPT models on hundreds of thousands of articles from USA TODAY and 18 local papers without permission, and that ChatGPT now summarises that journalism in a way that replaces reading the original. It covers every OpenAI model from GPT-1 to GPT-6.1.
Which papers are suing?
Nineteen titles owned by USA TODAY Co., including the Detroit Free Press, The Arizona Republic, the Indianapolis Star, the Milwaukee Journal Sentinel, The Tennessean, The Columbus Dispatch, The Des Moines Register and The Palm Beach Post. The case, USA Today Co. v. OpenAI Foundation, was filed in the Southern District of New York by attorney Steven Lieberman of Rothwell Figg. It names seven OpenAI entities and asks for a jury trial.
What does the complaint actually claim?
Three main things. First, copying: it says the papers’ content makes up more than 160,000 entries in WebText, the dataset used to train GPT-2, including 83,266 from usatoday.com, and more than 122 million tokens in C4, a filtered web dataset. Second, that OpenAI used tools to strip out copyright information attached to the articles. Third, and newest, that OpenAI deliberately trained its models to summarise articles instead of linking to them.
That third point is the heart of it. The filing includes examples where GPT-5.6, asked about a specific article by title, produced long summaries following the original’s structure. It quotes OpenAI’s Head of ChatGPT saying that once ChatGPT answers, there is "no good reason to click" through to the source. If nobody clicks, nobody subscribes, and that is the business damage the papers are claiming.
Where does the 250 million dollar figure come from?
US copyright law allows up to 150,000 dollars per wilfully infringed work, plus up to 25,000 per violation for removing copyright information. Across hundreds of thousands of articles, the theoretical ceiling is far above 250 million. The headline number is a floor the plaintiffs are asking for, not a cap, and actual damages in these cases are rarely close to the statutory maximum.
Does this matter more than the earlier lawsuits?
It adds weight rather than changing the picture. The paper asked for the case to be linked to the consolidated OpenAI copyright litigation already in the same court, which includes earlier suits from other publishers. It cites internal material including a claim that Microsoft supplied OpenAI with its Bing search index under a project codenamed Taxi. None of those allegations have been tested in court, and OpenAI has generally argued that training on public data is fair use.
The timing is awkward for OpenAI, which is already facing a California subpoena over its agents and pressure from the new White House incident rules. It also echoes the music industry’s case covered in our report on labels suing Anthropic. The common thread: whoever made the content AI models learn from wants paying, and the "no reason to click" argument is the clearest version yet of why.
Read Unite.AI’s breakdown of the complaint.
