Seattle Times and Newsday sue OpenAI and Microsoft, asking court to destroy AI models trained on their journalism

What happened

On September 4, 2026, The Seattle Times and Newsday filed a federal copyright infringement lawsuit against OpenAI and Microsoft in the U.S. District Court for the Southern District of New York (case 1:26-cv-07644). The 38-page complaint names nine OpenAI entities and Microsoft Corporation, alleging the companies scraped articles from the newspapers' websites — including content behind paywalls — and incorporated them into datasets used to train and operate ChatGPT, Microsoft Copilot, and Bing's AI features. The complaint cites an example in which an AI model reproduced 88 consecutive words from a Seattle Times article after being given its headline and URL. The claims include direct and vicarious copyright infringement, Digital Millennium Copyright Act violations, and trademark dilution.

The publishers seek damages and, unusually, an injunction requiring the destruction of all infringing copies of their work held in training datasets and AI models. Microsoft said it was surprised by the lawsuit and was willing to discuss solutions; OpenAI defended training on publicly available material as fair use. The case follows a similar 2023 suit by The New York Times against the same defendants.

Why it matters to creators

This is the first publisher complaint in this wave to make the destruction of AI models themselves a central remedy request: if granted, courts could order OpenAI and Microsoft to delete models and datasets containing the plaintiffs' work. The paywall-scraping allegation and the 88-word verbatim example (given only a headline and URL as a prompt) are concrete evidence on output reproduction, which is directly relevant to any writer or publisher weighing similar claims.

What is not yet known

No court has ruled on the merits; these are plaintiffs' allegations. Whether the destruction remedy — which copyright law permits in principle — will actually be applied to trained AI models remains untested, and the fair-use outcome is complicated by the U.S. Department of Justice's September 2, 2026 brief supporting OpenAI's fair-use defense (reported separately).