Before this filing, publishers suing OpenAI and Microsoft were mostly negotiating over money — licensing fees, damages, back payments for scraped archives. After it, at least two newsrooms are asking a federal judge to order something far more disruptive: the actual destruction of training datasets and AI models that incorporate their journalism. That shift, filed by the Seattle Times and Newsday in the Southern District of New York on September 4, 2026, changes the stakes of this litigation from a pricing dispute into a question of whether courts can order AI companies to unbuild parts of their systems.
Asking for a licensing fee is a negotiation; asking a judge to order a model destroyed is a demand that the defendant prove it can even comply.
The complaint also layers in a trademark dilution claim, arguing that chatbots have generated fabricated content falsely attributed to the two papers — a reputational harm distinct from lost clicks. That matters for the industry because it opens a second legal theory beyond copyright, one that doesn’t hinge on proving verbatim reproduction and could be harder for OpenAI to dismiss under a pure fair-use defense. The suit also cites industry data, reported by the Seattle Times, showing search referral traffic to midsize publishers fell 47% year over year in December 2025 — the kind of concrete economic-harm figure that plaintiffs’ lawyers will lean on to argue AI answers are substituting for subscriptions, not just competing with them.
The political backdrop makes this messier than a standard IP fight. Just last week, according to Mashable, the Trump administration filed a Statement of Interest in the parallel New York Times case arguing that a publisher win would threaten national security and undercut small newsrooms’ ability to compete using AI — the opposite of what the Seattle Times and Newsday are now arguing in their own filing. Add the disclosure that Microsoft Philanthropies has funded Seattle Times journalism projects and helped bankroll a $10 million Lenfest Institute AI fellowship that included both plaintiff newsrooms, and the case becomes a study in how tangled publisher-platform money has gotten even as the parties sue each other.
For data companies, the practical read is that licensing is looking cheaper than litigation risk. OpenAI has already cut deals with Axel Springer and the Associated Press rather than gamble on court outcomes, and Anthropic’s $1.5 billion settlement with authors and publishers, approved in July 2026, shows what a losing hand can cost. With roughly 400 newspapers having joined a separate suit in the same district back in June 2026, and Microsoft’s own roughly $13 billion stake in OpenAI tying it to every one of these outcomes, the incentive to settle rather than test the destruction remedy in front of a judge is only getting stronger. Watch whether the Southern District consolidates these cases with the Times litigation, and whether OpenAI responds to the Seattle Times and Newsday filing with a settlement offer rather than a fair-use defense.
We feel strongly that we must defend our content — which we spend millions of dollars a year to produce — from being used without our consent or compensation,
