Publishing Companies Sue Microsoft and OpenAI Over GPT Model Copyright Infringement

Six publishing companies have filed a lawsuit against Microsoft and ten OpenAI entities, alleging copyright infringement and the removal of copyright management information in the development of GPT models. The complaint, filed on October 2, 2026, in the U.S. District Court for the Southern District of Mississippi, seeks an injunction to remove all copies of the plaintiffs' registered works from the models and training sets.

The plaintiffs include Emmerich Newspapers, Ojai Media, and several Coopwood companies, which publish various newspapers and magazines. The defendants include Microsoft and multiple OpenAI entities, described as a web of interrelated companies. The lawsuit alleges that the defendants systematically crawled the plaintiffs' websites, including content behind paywalls, and copied articles onto their servers repeatedly as the models were updated.

The complaint claims that the models trained in this manner exhibit memorization, encoding retrievable copies of the training works. It also alleges that products like ChatGPT Search and Deep Research retrieve content from the plaintiffs' websites in real-time, incorporating it into responses. The plaintiffs argue that the defendants stripped author credits, publication names, copyright notices, and terms of use information from the articles.

The lawsuit includes three counts: direct copyright infringement, removal of copyright management information, and willful theft of the articles. The plaintiffs cite OpenAI's own statements and previous cases to support their claims. The case is assigned to Judge Henry Travillion Wingate, with the docket last updated on October 5, 2026.