OpenAI faces copyright lawsuits from major publications over AI training

America’s newspapers have found a common enemy, and it isn’t each other for once. The New York Times, the Chicago Tribune and now USA Today have all taken OpenAI to court over how ChatGPT learned to write.

The publishers allege that OpenAI fed their journalism into its models without permission, in ways that break copyright law. OpenAI says it didn’t need permission at all. And the US government has just picked a side.

The newest lawsuit carries a nine-figure price tag

The latest entrant is USA Today Co., which filed suit on October 8, 2026. It named both OpenAI and Microsoft as defendants.

The complaint alleges that the companies scraped a large volume of articles from 19 affiliated publications without authorization. USA Today Co. is seeking over $250 million in damages.

The New York Times got there first. It sued OpenAI and Microsoft in December 2023, and that case has since been folded into a larger multidistrict litigation, or MDL.

An MDL works a bit like a carpool for lawsuits. Instead of dozens of separate cases crawling through different courtrooms, related claims get consolidated so pretrial issues can be handled once. Numerous other publishers have joined that consolidated effort.

The process has not been gentle on OpenAI. Discovery disputes have produced orders requiring the company to hand over millions of ChatGPT logs, along with other internal documents.

Fair use is the whole ballgame

OpenAI’s defense rests on a single legal concept: fair use. The doctrine lets people and companies use copyrighted work without a license in limited circumstances.

The core dispute turns on whether that training is “transformative.” Courts look favorably on uses that create something meaningfully different from the original.

OpenAI maintains that its models do something fundamentally different from the articles they learned from. It also argues that its products don’t directly compete with the publishers in the marketplace.

The publishers see it the other way. They argue that AI firms reproduce and compete with their content without paying for it, eating into the revenue that funds the reporting in the first place.

Open motions for summary judgment on fair use are now in play. Summary judgment lets a judge resolve a case, or part of it, without a full trial when the key facts aren’t in dispute. Discussions on that front took place around September 4, 2026.

Washington weighs in for OpenAI

The most consequential development may not have come from either litigant. On September 2, 2026, the US Department of Justice intervened in support of OpenAI’s position.

The department filed a brief backing the company’s expansive reading of fair use. It argued that training large language models on copyrighted material falls within the doctrine.

The government framed the issue in terms of national interest, tying it to America’s standing in AI and technology.

What this means for AI companies and publishers

The stakes extend far past OpenAI. Nearly every major AI model depends on enormous volumes of text, and much of that text is copyrighted.

If OpenAI prevails on fair use, the current approach to building AI largely survives intact. Companies could keep training on publicly available material without negotiating licenses for every source.

If OpenAI loses, the calculus changes. A ruling that gives publishers tighter control over their content could force AI firms to rethink how they gather data.

Microsoft’s presence as a co-defendant matters here. As a publicly traded company tied closely to OpenAI, it gives investors a direct line of exposure to the outcome.

The things to watch are the summary judgment rulings on fair use and how courts handle the discovery fights over ChatGPT logs.