Seattle Times and Newsday sue OpenAI and Microsoft for infringement

The Seattle Times and Newsday have sued OpenAI and Microsoft for copyright infringement, alleging their journalism was used as training data without permission and that the companies' systems often reproduce passages from their reporting in answers to user queries. Microsoft is named because Copilot is built on OpenAI's technology.

The filing sits in a crowded docket already occupied by The New York Times, Ziff Davis, Merriam-Webster, and Encyclopedia Britannica. The two papers also join nearly 400 local newspapers that recently sued the same defendants, arguing chatbots siphon visits and subscription revenue by answering questions that once required a trip to the publisher's site.

Beyond money, the complaint asks for destruction of any copies of their works the companies hold, plus training datasets and AI models that incorporate them—a remedy that, if entertained, would raise the stakes well above a licensing negotiation.

This is less a novel legal theory than another brick on a wall already being built. The training-without-permission claim and the output-reproduction claim are the same twin arguments running through the Times and other publisher suits. What changes with each new plaintiff is leverage and venue pressure, not the core dispute.

Signal Desk's read: the destruction ask is the sharp edge. Damages and licensing can be priced; an order to destroy training sets and models that "incorporate" a publisher's work is existential for any lab whose corpus is entangled with news archives. Courts rarely grant that kind of relief lightly, but pleading it signals how far publishers are willing to push the remedy conversation while settlement talks proceed elsewhere.

The under-stated point is the local-paper coalition. Nearly 400 titles arguing lost subscription revenue frame AI answers as a direct substitute for reporting, not merely an upstream training issue. That theory, if it sticks, matters as much for ad-supported and metro newsrooms as for national brands with licensing desks.

OpenAI and Microsoft did not immediately respond. Expect the usual pattern: more plaintiffs, more discovery fights, and private licensing discussions that never match the complaint's maximalist remedy language. The likelier near-term outcome is accumulation of pressure around the Times docket, not a sudden judicial order wiping models.

Publisher copyright suits against OpenAI have proliferated since The New York Times' case, with Microsoft routinely named wherever Copilot is in the product mix. The Verge notes OpenAI began rolling out GPT-6 Astra around September 4, 2026—timing that puts a fresh model launch next to another wave of infringement claims, even if the suit's allegations concern training and outputs rather than that specific release.