The Seattle Times Sues OpenAI for Copyright Infringement

On September 4, 2026, The Seattle Times and Newsday filed a lawsuit against OpenAI and Microsoft in the U.S. District Court for the Southern District of New York. The complaint alleges that OpenAI and Microsoft "scraped the newspapers' website content, including articles accessible only through paid subscriptions," and incorporated those articles into training datasets used to train and operate ChatGPT, Microsoft Copilot, and Bing's AI features. The plaintiffs further allege that the foregoing AI products are capable of reproducing lengthy excerpts from their reporting, closely paraphrasing articles, and providing answers to users, thereby reducing the need for users to visit their websites or purchase subscriptions.

The complaint cites an illustrative example: when only the headline and URL of a Pulitzer Prize-winning article about the Boeing 737 MAX crash were entered, ChatGPT reproduced verbatim a passage of 88 words from that article. The plaintiffs contend that this fact directly demonstrates the AI model's ability to reproduce verbatim copyrighted news content.

The two newspapers seek an order requiring the destruction of copies of their works, as well as the training datasets or AI models containing those works. The plaintiffs also seek damages. The complaint further alleges that OpenAI diluted their trademarks by generating content falsely attributed to the plaintiffs.

Positions of the Parties and Legal Defenses

Plaintiffs' Position. In a letter to employees, Alan Fisco, President and CEO of The Seattle Times, stated: "We strongly believe we must defend the content we spend millions of dollars a year to produce from being used without our consent or payment to us." He also emphasized that the lawsuit "is not about impeding AI innovation," but rather ensuring that "innovation does not come at the expense of our business model."

Defendants' Response. An OpenAI spokesperson said in an email statement that its models are "trained on publicly available data and grounded in fair use," but declined to comment specifically on this lawsuit. A Microsoft spokesperson said in an email: "While we were surprised by this lawsuit, we understand the importance of local news, and we are always willing to sit down and discuss ways to resolve such disputes."

Analysis of the Core Legal Issues

The core legal issue in this case is whether the reproduction and use of copyrighted works in AI model training constitutes fair use, and how the substitutive impact of AI-generated content on the market for the original works should be determined.

(1) The Four-Factor Fair Use Analysis Framework

The four fair use factors set forth in Section 107 of the U.S. Copyright Act—the purpose and character of the use, the nature of the copyrighted work, the amount and substantiality of the portion used, and the effect of the use upon the potential market for or value of the copyrighted work—provide the core analytical framework for this case, with the effect on the potential market being central. The plaintiffs argue that OpenAI's and Microsoft's use does not constitute fair use because the AI products can reproduce original text and directly compete with news organizations for readers and subscription revenue, thereby constituting a substantial substitute for the market of the original works.

The defendants argue that AI training constitutes "highly transformative use." This position has received support from the U.S. Department of Justice. On September 2, 2026, the DOJ filed a statement of interest urging the court to adopt OpenAI's fair use defense, arguing that AI training "generally constitutes fair use of copyrighted material" and describing AI training as "highly transformative." The DOJ further stated that restricting the development of large language models would "severely hamper the progress of science and useful arts" as provided in the Constitution.

(2) Key Disputes in Discovery

In the related case (The New York Times v. OpenAI and Microsoft), the parties are engaged in discovery and summary judgment motions. Microsoft provided an expert retained by the publishers with data on 8.2 million Copilot chat logs, arguing that only 59,545 logs contained at least 16 words matching news content, and only 24 responses displayed 30 or more matching words, thereby demonstrating the transformative nature of its use.

In addition, news organizations have separately filed a motion for sanctions against OpenAI, alleging spoliation of evidence. That motion remains pending before the court. China Intellectual Property Lawyer Network will continue to follow subsequent developments.