Seattle Times and Newsday sue OpenAI and Microsoft, seek destruction of AI models trained on their reporting

Two more publishers sue OpenAI and Microsoft

The Seattle Times and Newsday have filed a copyright infringement lawsuit against OpenAI and Microsoft, according to a September 6 report from The Verge. The two publishers allege that OpenAI used their journalism as training data for its AI models without permission, and that its systems often reproduce passages of their reporting when answering user queries.

Microsoft is named alongside OpenAI because Copilot, Microsoft's assistant product, is built on OpenAI's technology. The Verge reports that neither company immediately responded to a request for comment.

The publishers want the models destroyed, not just compensation

The most consequential part of the filing is the remedy being sought. Beyond damages, the Seattle Times and Newsday are asking the court to order the destruction of any copies of their works held by the two companies, as well as the training datasets and AI models that incorporate their material.

That request targets the trained systems themselves rather than just the underlying data. Because a neural network does not store individual articles in a form that can be surgically removed, a destruction order of this kind would in practice mean rebuilding or retiring systems that already sit behind shipped products. It frames the alleged infringement as baked into the technology rather than a licensing gap that could be closed with a payment and a content takedown.

A widening wave of publisher litigation

The new suit is not an isolated action. The Verge notes that The New York Times, Ziff Davis, Merriam-Webster and Encyclopedia Britannica have all brought similar lawsuits against OpenAI over its use of their content in training data.

The two outlets also join a much larger group: nearly 400 local newspapers recently sued OpenAI and Microsoft, arguing that chatbots which answer questions using publishers' reporting reduce the need for readers to visit their sites, costing the outlets valuable subscription revenue. By adding the Seattle Times and Newsday, the roster of plaintiffs now spans national, regional and local newsrooms alongside dictionary and encyclopedia publishers.

Why it matters

Demanding the destruction of trained models raises the stakes well beyond a typical copyright dispute. If a court ever granted such a remedy, OpenAI and Microsoft would face the prospect of retraining or withdrawing systems that power products like Copilot, and the ruling would cast a shadow over every model built on large-scale web-scraped text.

For publishers, the underlying argument is economic: if chatbots hand out answers drawn from their reporting, readers have less reason to subscribe, and the funding model that pays for original journalism erodes. With the New York Times, hundreds of local papers and several reference publishers all now in court against OpenAI, the combined outcome of these cases could shape how AI training data is sourced, licensed and paid for across the industry.