The Seattle Times and Newsday have jointly sued OpenAI and Microsoft over AI training on copyrighted journalism. Engadget reported the filing. The papers call generative AI a snake eating its own tail that could destroy the newsrooms that feed it. That line is dramatic, but the business claim is straightforward. They say the companies scraped articles in ways that bypass paywalls, then used that material to train models that compete with the original reporting.
According to the lawsuit, AI products give readers a generated alternative to news articles and cut traffic and digital ad revenue for the publishers. The complaint also says the models can hallucinate false information and attribute it to the news outlets, and that copyright management information was removed from articles. Those are the kinds of details that turn a general training fight into a concrete product complaint.
This is not the first case of its kind. The New York Times sued OpenAI and Microsoft in 2023. CNN sued Perplexity earlier this year. Other outlets have taken the licensing route instead. AP and Vox Media partnered with OpenAI so the company can train on their material under contract. The Seattle Times case stands out for another reason. Microsoft and OpenAI have funded some of the paper's journalism projects and fellowships. A Microsoft spokesperson told GeekWire the company is surprised by the lawsuit and is always happy to sit down and explore solutions.
Why should someone who just reads the news on their phone care? Because the fight is really about who pays for trustworthy reporting in an AI world. If models can summarize paywalled work without a deal, the incentives to fund reporters get weaker. If courts push companies toward licenses, you may see more credited answers, clearer sourcing, and fewer mystery citations. Either outcome changes what shows up in your chat window.
I write software for a living, and I use these models every day. I also know newsrooms are not optional infrastructure. Good local reporting is how people learn about schools, courts, and storms. When a hometown paper and a major regional title say AI training hollowed out their traffic, that is a product-design problem for the whole industry, not only a courtroom drama.
There is a practical takeaway for builders and readers. Prefer tools that show sources. Support publishers you rely on with subscriptions when you can. If you ship an AI feature that answers from news, budget for licenses or stick to public summaries with clear links. Microsoft saying it wants to talk is useful. The lawsuit says talking alone has not been enough for these two papers.
This case will not settle the whole copyright war overnight. It does add pressure in a week when OpenAI is already under scrutiny for agent safety. For regular folks, the hopeful path is boring and healthy. Better licensing. Clearer attribution. Models that point you back to the humans who did the reporting. That is how AI stays useful without starving the sources it learns from.
