Source: Law Commentary

Seattle Times and Newsday sued OpenAI and Microsoft, alleging the companies copied hundreds of thousands of their articles without permission to train AI systems and asking a federal court to destroy models and training datasets that incorporate their journalism. The complaint, filed September 4 in the U.S. District Court for the Southern District of New York, targets uses of the publishers' work in ChatGPT, Microsoft Copilot and AI features in Bing and says stories were scraped from their websites, including articles behind paywalls. The filing says ChatGPT reproduced 88 consecutive words from one Seattle Times article after being provided that article’s headline and URL and cites the paper’s Pulitzer Prize-winning Boeing 737 MAX coverage as an instance of reporting that can surface in AI-generated responses.

The newspapers seek damages and an injunction and specifically ask the court to apply federal copyright remedies — including destruction of infringing copies — to copies of their journalism, associated training datasets and AI models containing their work, a relief that raises a novel question about whether courts can order destruction of trained large language models. No destruction order has been entered; the publishers must prove their claims and persuade the court such relief is warranted, with fair use expected to be central. OpenAI contends training on publicly available material is transformative and protected by fair use, while Microsoft said it was surprised and is open to discussion. The complaint notes a prior 2024 Lenfest Institute AI fellowship supported by OpenAI and Microsoft was not a blanket license to use the publishers’ work, and Seattle Times CEO Alan Fisco described filing the suit as a difficult decision to protect journalism the company funds.