The Pulse
USA Today Publishers Sue OpenAI Over Training and ChatGPT Outputs
USA Today Co. and publishers of 19 publications sued OpenAI in New York on October 8, 2026. They allege the company used their journalism to train its models and repackaged articles in chatbot responses, and seek more than $250 million.
AI.info Team · Written with AI ·
“OpenAI's internal communications confirm that it deliberately targeted news content and knew its models were built to reproduce it.”
USA Today, in a statement quoted by MLex
USA Today Co. and publishers of 19 publications filed a copyright lawsuit against OpenAI on October 8, 2026, accusing the company of using their journalism to train its models and repackaging that work in responses to ChatGPT users. The complaint was filed in the U.S. District Court for the Southern District of New York. It seeks more than $250 million in damages and a court order barring the alleged infringement.
The publishers’ claims remain allegations, not findings. The case puts two connected uses of news content in dispute: its inclusion in material used to train AI systems, and the appearance of article content in answers delivered to users.
19 publications join the complaint
The plaintiffs say their copyrights cover content from 19 publications, including USA TODAY, The Arizona Republic, the Detroit Free Press and the Indy Star. The group also includes local titles such as The Tennessean and The Columbus Dispatch, extending the suit beyond the national newspaper to the company’s regional news portfolio.
USA Today Co. alleges that OpenAI copied hundreds of thousands of articles and other materials without permission. Its complaint argues that the use was not limited to training: the publishers say OpenAI also repackaged their reporting in outputs for ChatGPT users. They characterize those answers as a threat to the value of their original reporting and the audience it draws.
Publishers challenge both training and answers
The complaint’s focus on chatbot outputs gives the dispute a dimension beyond whether copyrighted material can be used to train a model. The publishers allege that generated responses can deliver news content in a form that substitutes for reading the original article. That raises questions about how much material a system returns, whether its answers direct users back to publishers, and what commercial effect those answers have.
Those questions sit alongside the training dispute. The plaintiffs contend OpenAI used their material without a license; the company has not publicly answered this complaint in the initial coverage of the filing. The lawsuit does not resolve whether the alleged uses infringe copyright or qualify for a legal defense.
$250 million in damages and a request to block the use
The publishers seek damages exceeding $250 million and ask the court to stop the conduct they describe as infringement. OpenAI’s response will help define how it contests the claims, but the case will also turn on the publishers’ evidence about what material was used, how it entered model development, and what ChatGPT returns to users.
The suit joins other publishers’ legal challenges to AI companies’ use of copyrighted work. Technology companies have argued in such disputes that training can constitute fair use because it transforms source material into something new; the legal question depends on the particular facts and uses at issue.
For now, no court has ruled on USA Today Co.’s claims. The immediate dispute is whether the publishers can show that OpenAI’s alleged training and user-facing outputs crossed the line from using news content to infringing it.