The Pulse
Seattle Times and Newsday Sue OpenAI and Microsoft Over AI Training
The Seattle Times and Newsday filed a federal lawsuit accusing OpenAI and Microsoft of copying their journalism to train and operate AI products. The publishers seek damages and an order covering allegedly infringing copies, datasets and mo
AI.info Team ·
When a user supplied ChatGPT with only the headline and web address of a Seattle Times investigation into the Boeing 737 MAX crashes, the chatbot allegedly returned an 88-word passage from the article nearly verbatim. The example appears in a federal complaint filed September 4 by The Seattle Times Company and Newsday against OpenAI, Microsoft and nine OpenAI corporate entities.
The newspapers accuse the companies of copying their reporting without permission, using it in datasets and models behind products including ChatGPT, Copilot and Bing Chat, and generating answers that compete with the original publications. The complaint demands a jury trial, financial relief and an order requiring the destruction or impoundment of allegedly infringing copies, training datasets and models.
OpenAI has not publicly responded to the filing in the sources reviewed for this article. A Microsoft spokesperson told GeekWire that the company was “surprised by the lawsuit” but was “always happy to sit down and explore solutions to this type of dispute.”
The 88-word example at the center of the complaint
The case, filed in the U.S. District Court for the Southern District of New York, alleges that OpenAI and Microsoft obtained large quantities of Seattle Times and Newsday journalism through automated scraping. The publishers say the companies bypassed paywalls, ignored website terms and incorporated copies into datasets derived from WebText, WebText2, Common Crawl and Microsoft’s Bing search index.
Those assertions are allegations, not findings by the court. The complaint says the copied material was used to train, fine-tune and “ground” large language models, while also claiming that the companies stored copies for retrieval-augmented generation and for answering user queries.
The Boeing 737 MAX example is the complaint’s clearest illustration of the publishers’ argument about output. According to the filing, a prompt containing the article’s headline and URL produced an 88-word sequence from Seattle Times coverage of the crashes. The newspaper says the example shows that the dispute is not limited to the abstract question of whether training on copyrighted works is lawful; it also concerns what commercial AI systems return to users.
The complaint argues that ChatGPT and Microsoft’s products do more than analyze or summarize journalism. It says they can reproduce, closely paraphrase and synthesize reporting into answers that substitute for visits to seattletimes.com or newsday.com, competing for readers, advertising and subscription revenue.
A joint case with seven legal claims
The docket identifies The Seattle Times Company and Newsday LLC as the plaintiffs in case number 1:26-cv-07644. The defendants are Microsoft Corporation and OpenAI, Inc., OpenAI GP, LLC, OpenAI, LLC, OpenAI OpCo, LLC, OpenAI Global LLC, OAI Corporation, OpenAI Holdings, LLC, OpenAI Foundation and OpenAI Group PBC.
The complaint contains seven counts. They include direct copyright infringement, vicarious copyright infringement, two claims under the Digital Millennium Copyright Act concerning the alleged removal and distribution of copyright-management information, and trademark-dilution claims under federal, Washington and New York law.
The trademark claims give the newspapers a second route beyond the training dispute. The publishers allege that AI products can associate their names with generated material they did not publish, including inaccurate or fabricated content. They say that association can weaken the value of the Seattle Times and Newsday marks even when an output does not reproduce a complete article.
The filing identifies registered works belonging to both newspapers in exhibits attached to the complaint. It does not state a single dollar figure for damages. Instead, the publishers seek actual damages, profits, statutory relief where available, injunctive relief, attorneys’ fees and other remedies allowed by law.
Microsoft is being sued by its hometown newspaper
The Seattle Times case carries an unusual corporate relationship. Microsoft is headquartered in Redmond, Washington, while The Seattle Times is the region’s largest major daily newspaper and has covered the company for decades.
Microsoft Philanthropies has supported some Seattle Times journalism projects, and Microsoft and OpenAI jointly helped fund a $10 million artificial-intelligence fellowship run by the Lenfest Institute in 2024. The fellowship included The Seattle Times and Newsday among its first participating newsrooms. The newspaper has said those relationships do not compromise its editorial independence.
Alan Fisco, president and chief executive of The Seattle Times Company, described the decision in a memo quoted by GeekWire: “This was not an easy decision. However, 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.”
The Seattle Times Union, which represents more than 160 newspaper employees, said it supported the lawsuit. The union also criticized the company during contract negotiations for refusing to guarantee that it would not replace non-reporter newsroom jobs with AI, placing the legal action alongside an internal dispute over how the newspaper itself will use the technology.
Two regional publishers challenge the economics of AI answers
The filing presents both newspapers as businesses that depend on paying readers and direct relationships with their audiences. The Seattle Times says its website receives about 6 million monthly unique visitors and 29 million monthly pageviews, with roughly 108,000 digital-only subscribers and about 100,000 print subscribers. Newsday says it has more than 60,000 digital subscribers, more than 60,000 daily print copies and about 51 million average monthly pageviews.
Both publishers operate significant digital paywalls. The Seattle Times says readers receive one free pageview before payment is required. Newsday says it has maintained a complete hard gate for its editorial text since August 2022.
The complaint argues that those subscription systems make unlicensed copying particularly damaging. A reader who receives the substance of a reported story from a chatbot may have less reason to open the publisher’s page, encounter its advertising or pay for access. The publishers describe AI-generated responses as competing informational products rather than as a conventional search index that directs users to the original source.
That argument is central to the fair-use fight likely to follow. OpenAI and Microsoft have argued in other litigation that using copyrighted material to train models can serve a different, transformative purpose. The Seattle Times and Newsday answer that the products reproduce or closely mimic the same expressive work and deliver the same information to the same potential audience.
The filing joins a larger fight in Manhattan
The new case arrives as other news organizations pursue related claims against OpenAI and Microsoft in the same federal district. The New York Times filed its lawsuit in 2023, while the New York Daily News, Ziff Davis and the Center for Investigative Reporting are among the publishers involved in related proceedings consolidated before U.S. District Judge Sidney H. Stein.
On September 4, the parties in that consolidated litigation also filed motions for summary judgment. A ruling in that case could shape how courts analyze the use of news articles in model training, although it would not automatically decide every factual or legal issue raised by the Seattle Times and Newsday.
The Justice Department recently filed a brief supporting arguments from OpenAI and Microsoft in the New York Times case, warning that a ruling for the publishers could hinder U.S. AI development. The new complaint puts the publishers’ response in direct terms: they say the companies built valuable products by copying the material that news organizations spend millions of dollars producing, then used those products to compete for the same audiences.
OpenAI has signed licensing agreements with several news organizations, including the Associated Press, News Corp and Axel Springer. The Seattle Times complaint says publicly disclosed terms from three such agreements total more than $300 million, while arguing that licensing selected publishers does not resolve the alleged use of its own and Newsday’s work.
For now, no judge has ordered OpenAI or Microsoft to destroy a model or dataset, and the complaint has not been tested through discovery or trial. The immediate legal question is whether the two newspapers can move their allegations about scraping, training, retrieval and generated output into a case that survives the defendants’ first motions. The answer will begin with the 38-page complaint filed in Manhattan on September 4, 2026.