The Seattle Times and Newsday sued OpenAI and Microsoft this week, becoming the latest news organizations to accuse the companies of training AI models on their journalism without permission, according to TechCrunch and The Verge.
The complaint alleges that OpenAI used the two outlets' reporting as training data and that its chatbots reproduce passages from that reporting when users ask about news events, according to The Verge. Microsoft was named as a co-defendant because its Copilot assistant runs on OpenAI's technology, according to The Verge.
Training data is the text a company feeds a model so it learns to generate answers on its own. When that text is copyrighted journalism collected without a license, the publishers argue they are funding a product that competes directly with them for readers.
The lawsuit describes generative AI as "a snake eating its own tail," arguing it could "destroy the very organizations" that produced the material it was trained on, according to TechCrunch. It also calls the companies "rapacious consumers" that deliver back "copies and derivative imitations" of the journalism they consumed, according to TechCrunch.
The complaint warns that, left unchecked, this cycle could leave the journalism industry "broken beyond repair," according to TechCrunch, since fewer readers click through to the original reporting once a chatbot can summarize it directly.
A legal front that keeps widening
The Seattle Times and Newsday follow The New York Times, which sued OpenAI and Microsoft over copyright infringement back in 2023 and whose case is still being litigated, according to TechCrunch and The Verge. Other publishers have joined since, including Ziff Davis, Merriam-Webster and Encyclopedia Britannica, according to The Verge.
The Verge reports that close to 400 local newspapers have also sued OpenAI and Microsoft, arguing that AI chatbots reduce the need for readers to visit their sites directly and cut into subscription revenue as a result.
What the newest suit is asking for
The Seattle Times and Newsday are seeking the destruction of any copies of their work held by OpenAI and Microsoft, along with the training datasets and AI models built from that material, according to The Verge.
TechCrunch notes that the Seattle Times' case stands out because Microsoft and OpenAI have previously funded some of the newspaper's own journalism projects and fellowships, a relationship the lawsuit does not erase.
A Microsoft spokesperson told GeekWire the company was "surprised by the lawsuit" but remained "always happy to sit down and explore solutions to this type of dispute," according to TechCrunch. OpenAI and Microsoft did not immediately respond to a separate request for comment from The Verge.
Why it matters beyond the newsroom
Every new suit adds pressure on OpenAI and Microsoft to negotiate licensing deals with publishers instead of litigating each claim individually. The pattern is already familiar from the music industry, where labels have pursued similar copyright cases against AI companies over training data.
At the center of every one of these cases sits the same unresolved legal question: whether training a model on copyrighted text without a license counts as fair use under US copyright law. Courts have not settled that question, which is why publishers keep filing suits rather than waiting for a single precedent to decide the issue for them.
The outcome will help determine whether news organizations get paid when their reporting is used to train the same models that increasingly answer readers' questions directly, cutting out the click to the original story.
For now, the case joins a crowded docket that already includes some of the largest newspaper publishers in the country, and no court ruling has been issued in either direction.
