Seattle Times and Newsday sue OpenAI and Microsoft over AI training

This tension reached a legal breaking point on Friday, September 5, 2026, when The Seattle Times and Newsday filed a joint lawsuit against OpenAI and Microsoft in the U.S. District Court for the Southern District of New York.

The news organizations allege that the tech giants did not merely scrape public data, but methodically bypassed paywalls to ingest content intended for paying subscribers. By incorporating this journalism into datasets used to power ChatGPT, Microsoft Copilot, and Bing’s AI features, the publishers argue the companies have created a "snake eating its own tail" dynamic: AI systems rely on high-cost professional journalism to function, but the resulting products dilute the market for those same newsrooms, threatening to leave the industry "broken beyond repair."

Related YouTube video

Beyond Scraping: The Paywall Allegations

The publishers claim OpenAI and Microsoft bypassed technological restrictions designed to protect subscription-only content.

The lawsuit further alleges that these AI models frequently reproduce original reporting verbatim or closely paraphrase it, effectively allowing users to avoid visiting the source websites or purchasing subscriptions. Additionally, the newsrooms claim the models "hallucinate," attributing false information to the two outlets, while stripping away copyright management information from the original articles.

The stakes for the newsrooms are financial and operational. Alan Fisco, President and CEO of The Seattle Times, informed employees that the organization must defend content that costs millions of dollars a year to produce from being used without consent or compensation.

The Paradox of Partnership

The litigation is marked by a sharp reversal in business relations. According to reporting from Whalesbook, Microsoft and OpenAI previously collaborated with these media groups, including co-funding a $10 million AI fellowship program in 2024.

This divide is mirrored across the broader media landscape, where publishers have chosen two distinct paths in response to the rise of generative AI:

Competing Legal Frameworks: US vs. EU

In the United States, OpenAI has defended its practices by stating its models are trained on "publicly available data" and are grounded in the doctrine of fair use.

However, the legal landscape differs significantly in Europe. As reported by The Next Web, the EU’s general-purpose AI code—which OpenAI has signed—explicitly commits signatories not to circumvent access restrictions, specifically naming subscription models and paywalls. While this code has applied since August 2025 and does not reach conduct before that date, it reflects a stricter approach to access restrictions than currently exists in US law.

there is no European equivalent to the US "fair use" defense. A German court ruled in July that works stored within model parameters constitute infringement at the point of storage, a precedent that could provide European publishers with a more direct path to victory than their American counterparts.

Financial and Technical Implications

The plaintiffs are not seeking a simple settlement; they are demanding the "impoundment and/or destruction" of datasets and AI models that incorporated their journalism.

Microsoft expressed surprise at the lawsuit but stated it remains "happy to sit down and explore solutions," according to a spokesperson. The timing of the filing is particularly pointed, occurring on the same day Microsoft asked the same court to reject theories regarding lost licensing fees and market dilution in a separate matter.