The Seattle Times and Newsday have sued OpenAI and Microsoft, alleging the companies used their journalism without permission to train generative artificial intelligence systems. The lawsuits make the two newspapers the latest publishers to argue that large language models were built on copyrighted reporting.
At issue is whether ingesting news articles to improve chatbots and related commercial products requires a license. The papers contend that unauthorized copying of their work can substitute for original reporting and weaken the market for local and regional news.
Other U.S. news organizations have filed similar claims as generative tools moved from research projects into widely used products. The disputes often name both OpenAI, which develops the models, and Microsoft, which invests in and distributes the technology.
The lawsuits make the two newspapers the latest publishers to argue that large language models were built on copyrighted reporting.
AI developers have generally maintained that training on publicly available text can qualify as fair use and that their systems may send readers back to publishers. Courts have not yet issued a single ruling that settles how copyright law applies to model training at commercial scale.
For newsrooms under financial strain, the cases are as much about future licensing as they are about past use. A clearer legal standard would help determine whether publishers can charge for archival content used to train commercial AI, or whether that material remains freely available to developers.
