Two major American regional newspapers have launched a federal lawsuit against OpenAI and Microsoft, alleging the technology giants used decades of original reporting to train artificial intelligence models without permission or payment.
The Seattle Times and Newsday filed the joint 38-page complaint in the U.S. District Court for the Southern District of New York on Friday 4 September 2026. The legal action marks a significant escalation in the conflict between the news industry and AI developers over the “fair use” of intellectual property.
The publishers allege that OpenAI’s ChatGPT and Microsoft’s Copilot programmes were built by “scraping” millions of news articles, effectively creating a commercial product that competes directly with the original sources. The complaint includes specific evidence of the AI tools reproducing verbatim stretches of reporting, in one instance allegedly outputting an 88-word sequence identical to a published news story.
Alan Fisco, President and CEO of the Seattle Times, indicated that the legal action seeks not only financial damages but also the “impoundment and destruction” of existing AI datasets and models that incorporate the newspapers’ content.
US Government intervention
The filing follows a dramatic intervention by the US Department of Justice (DOJ) in a related case involving the New York Times. On 1 September 2026, the DOJ filed a Statement of Interest siding with the tech companies’ defence.
Federal lawyers argued that stalling the development of large language models would “threaten national security” and potentially aid “foreign adversaries” in the global race for AI dominance. The DOJ’s position suggests that the US government views the rapid advancement of AI as a strategic priority that may outweigh individual copyright claims.
This stance creates a sharp contrast with recent policy shifts in the United Kingdom. In March 2026, the UK government confirmed it would not pursue a broad copyright exception for AI training. Science Secretary Liz Kendall indicated at the time that the UK would instead prefer market-led licensing solutions to ensure creators are compensated.
According to a report and impact assessment on copyright and artificial intelligence, the UK authorities have moved away from earlier proposals that would have allowed AI firms more freedom to use copyrighted works for commercial research.
The ‘Fair Use’ debate
OpenAI has consistently maintained that training AI models on publicly available data constitutes “fair use” under US law. The company argues that its systems create transformative new works rather than simple copies of the data they ingest.
However, the newspapers describe the current AI ecosystem as a “snake eating its own tail.” The publishers argue that by diverting audiences and advertising revenue away from local newsrooms while using their content for free, the tech companies are destroying the very primary sources of information that the AI models rely on for accuracy.
The lawsuit in New York is currently in its early procedural stages. No date has been set for a hearing, and the court has not yet ruled on whether the case will proceed to a full trial. This development follows a series of similar legal challenges from authors, visual artists, and music publishers, all seeking to establish new legal boundaries for generative artificial intelligence.
