The Seattle Times and Newsday sued OpenAI and Microsoft on September 4 in Manhattan federal court, alleging copyright infringement and the stripping of copyright management information from their journalism. The filing landed in the same week the U.S. Department of Justice filed a twenty-page statement of interest in the consolidated publisher litigation arguing that training AI models on copyrighted work is "extraordinarily transformative" and protected by fair use. Publishers keep filing. The government just told the court whose side it is on.
The new complaint is notable less for its claims, which track the cases that came before it, than for the evidence it marshals and the remedy it seeks. The papers allege that OpenAI's models reproduced 88 consecutive words, verbatim, from the Seattle Times' Pulitzer Prize-winning Boeing 737 MAX coverage when prompted with a headline and URL, and that chatbot outputs substitute for visiting the publications and subscribing. The complaint calls generative AI "a snake eating its own tail" that could leave journalism "broken beyond repair," and it asks the court to order destruction of the training datasets and models that incorporated the papers' work.
What the two papers allege
The claims span copyright infringement, violations of the Digital Millennium Copyright Act for removing authors' names, titles, and notices from articles, and trademark dilution through false attribution when models credit fabricated statements to the publications. The suit names the datasets at issue, OpenAI's WebText corpora and Common Crawl, along with Microsoft's Bing index as a source of copies. It alleges scraping of paywalled and subscriber content in apparent violation of the sites' terms, and notes that Newsday's robots.txt explicitly barred OpenAI and Common Crawl from crawling its site.
The two papers are regional institutions with digital subscription businesses at the center of the harm claim. The Seattle Times has roughly 108,000 digital subscribers and Newsday about 60,000. The complaint cites traffic data showing search-referral traffic to mid-size dailies fell about 47 percent from December 2024 to December 2025, against roughly 22 percent for the largest national outlets, and argues the decline tracks AI answers displacing clicks. Seattle Times president Alan Fisco put the core position in one sentence: "We must defend our content from being used without our consent or compensation."
The licensing market is the complaint's structural argument. OpenAI has struck content deals with the Associated Press, Axel Springer, the Financial Times, The Atlantic, Vox, News Corp, and others, with three public deals reportedly worth more than $300 million combined. If a license market exists, the papers argue, then training without a license is not fair use; it is a refusal to pay in a market the defendant itself helped create. OpenAI's position, unchanged across every suit, is that its "models are trained on publicly available data and are grounded in fair use". Microsoft said it is "always happy to sit down and explore solutions to this type of dispute."
There is one wrinkle the filing cannot avoid. Microsoft Philanthropies underwrites some Seattle Times projects, and Microsoft and OpenAI co-funded a $10 million journalism fellowship involving both newsrooms. The Times asserts editorial independence, but the relationship is the kind of detail that will come up in discovery and in coverage alike.
The DOJ picks a side
The government's intervention, filed around September 2, is the more significant development for the litigation as a whole. In the New York Times case, the consolidated anchor of the publisher docket, the DOJ argued that training AI on copyrighted text is fair use, that liability would slow American AI development, and that the stakes include national security and economic competitiveness. Associate Attorney General Stanley Woodward called it "a historic statement of interest," which it is: the federal government has now weighed in on the defendant's side of the highest-stakes copyright litigation in a generation.
The timing with the new suit is coincidence, but the pairing is the story. Publishers see a shrinking referral economy and a growing licensing line item, and they are filing suit to move one into the other. The administration sees a technology race and has filed to keep the training pipeline open. Both cannot win on the fair use question the New York Times case will decide.
The hurdle the new suit faces
The consolidated litigation already contains a warning specific to the DMCA claims. In Raw Story and AlterNet's suit, the court dismissed claims that OpenAI removed copyright management information, holding the publishers lacked standing to sue, and later refused to let them amend. That precedent sits directly in the path of the new complaint's DMCA count, which will have to survive the same scrutiny.
The procedural landscape is otherwise set. The New York Times case, filed in December 2023, sits before Judge Sidney Stein in the Southern District of New York, with the publisher cases consolidated into a multidistrict litigation captioned In re OpenAI Copyright Litigation. Summary judgment motions landed in the same week as the new filing, with news organizations urging the court to reject the fair use defense. Around four hundred local newspapers, led by a group assembled around Alden Global Capital, are litigating separately. None of the major publisher suits has settled, a fact that surprises observers who expected the licensing deals to clear the docket.
The new case's strongest asset is not its legal novelty, which is modest, but its facts. Two regional papers, one of them with a Pulitzer-winning investigation reproduced word for word in a model's output, make a concrete record that broad fair use arguments have a harder time absorbing than they do against abstractions about training data. The government's brief argues the abstract case well. The complaint in this suit argues the concrete one, 88 words at a time.
Primary sources
- The Seattle Times' own story on the suit for the papers' position and the Fisco statement.
- TechCrunch for the complaint's structure, the Lenfest fellowship detail, and the relief sought.
- Reuters reporting, syndicated by the Star-Advertiser, for the parties' statements.
- USA Today for the DOJ's statement of interest and its fair use position.
- The Verge for the case's position within the consolidated litigation.