The facts
USA Today Co. and 13 Gannett‑owned publications filed a complaint in the Southern District of New York on October 8, 2026, seeking damages exceeding $250 million. The plaintiffs own copyrights in content from 19 newspapers, including USA Today, the Detroit Free Press, the Arizona Republic, the Des Moines Register and the Milwaukee Journal Sentinel. The complaint alleges OpenAI trained its large‑language models on that copyrighted material without permission, citing internal emails where Greg Brockman called the models “particularly good at predicting text of news articles.” It also references Sam Altman’s testimony before the British House of Lords that training today’s leading AI models would be impossible without copyrighted works. The suit requests a jury trial and asks the court to order the destruction of any OpenAI model or training set that incorporates the plaintiffs’ content, invoking 17 U.S.C. § 503(b). The complaint is signed by Steven Lieberman.
Our take
Well, shit. Gannett isn’t just asking for a few bucks; it’s demanding the court yank every OpenAI model that ever glanced at a story from its newspapers. That’s not a lawsuit, it’s a legal tantrum with a docket number, the equivalent of a toddler screaming for the TV to be smashed because a cartoon used a crayon they once owned. The complaint lovingly quotes OpenAI executives admitting they needed copyrighted material to build their product - Sam Altman told the British House of Lords it was “impossible” to train today’s AI without it, and Greg Brockman bragged the models were “excellent at news.” Then Gannett turns around and asks a judge to order the “destruction” of any model that incorporated that content. It’s like arresting a baker for using flour, getting the baker to confess that bread needs flour, and then demanding every loaf in the city be burned. The math is equally absurd: the plaintiffs calculate damages at up to $150,000 per willful infringement plus $25,000 per violation of copyright‑management information, yet the total demand tops out at just over $250 million - hardly enough to cover OpenAI’s coffee budget. The real punch line is the annihilation request. Gannett isn’t after compensation; it wants to see the machine that learned from its work unmade. That’s not protecting copyright, it’s a dying industry’s last, pathetic lunge at a technology that already outpaced it. They saw the future eat their lunch and responded by suing for the right to shove their fingers down its throat and make it vomit. Good fucking luck with that, you absolute clowns. The judge will probably use ChatGPT to draft the order dismissing your case.
