Seattle Times and Newsday Sue OpenAI and Microsoft

The Seattle Times and Newsday have sued OpenAI and Microsoft in federal court, arguing that both companies copied their journalism without permission to build and run AI products including ChatGPT, Microsoft Copilot and Bing's AI features. The complaint was filed on 4 September against nine OpenAI entities and Microsoft, and reported by TechCrunch and the Spokesman-Review.

The papers allege that the companies scraped their websites, including articles behind paywalls, and folded that work into the datasets used to train large language models. The suit puts the stakes in unusually stark language, describing generative AI as "a snake eating its own tail" and warning that the technology could "destroy the very organizations" whose reporting it depends on. Left unchecked, the complaint argues, the business of journalism could be "broken beyond repair."

What they want

The remedies sought go further than money. The publishers ask for copyright and DMCA damages and any attributable profits, but they also want permanent injunctions and, most pointedly, an order to impound or destroy the models and training datasets that contain their work. Deleting a trained model is a far heavier ask than a licensing cheque, and it is the part of these cases that most worries the labs.

OpenAI said its models are trained on publicly available data and that its use is grounded in fair use, the same defence it has raised since the New York Times sued it in 2023. Microsoft said it was "surprised by the lawsuit" but "always happy to sit down and explore solutions to this type of dispute."

A widening front

The filing adds two more mastheads to a growing line of publishers and rights holders testing whether training on copyrighted work counts as fair use. The same question sits behind the music industry's case against Anthropic, where Sony and Warner Chappell have sued over songs used to train Claude. Courts have yet to give a clear answer, and until they do, every frontier lab is building on contested ground.

There is an irony the newspapers lean on deliberately. If AI systems hollow out the outlets that produce reliable reporting, the models lose the very source of fresh, trustworthy text they need to stay useful. Whether that argument lands as law or merely as rhetoric is now for a judge to decide.