Seattle Times sues OpenAI and Microsoft over copyright infringement

The lawsuit now sits inside a national copyright battle

The Seattle Times’ copyright case against OpenAI and Microsoft turns on a direct question: can an AI developer, and the technology partner that supplies its infrastructure and products, use a newspaper’s journalism as training material without first obtaining a license? The newspaper says no. Its complaint accuses the companies of building valuable AI systems with reporting created through costly newsroom labor, while denying the publisher control, attribution and compensation for that work [1].

That claim is no longer merely a local dispute between a Seattle newsroom and two technology giants. During the 72-hour review window ending on September 5, 2026, the most important fresh development was the Justice Department’s intervention in the broader OpenAI copyright litigation. The department filed a statement of interest backing OpenAI’s argument that training large language models on copyrighted material found online is protected by fair use, a doctrine that permits certain unauthorized uses under U.S. copyright law [2].

The filing matters because the Seattle Times’ case rests on the same legal fault line. Publishers argue that AI companies are copying expressive works at industrial scale and using them to build substitutes for the original journalism. OpenAI and Microsoft argue that model training is transformative, because the system does not merely republish articles but analyzes patterns across enormous volumes of text to generate new responses [3].

What the government’s position changes

The Justice Department did not decide the case. It cannot. The court remains responsible for applying copyright law to the facts. But the government’s statement gives OpenAI and Microsoft a powerful policy argument at a moment when judges are being asked to define the rules for AI training across books, newspapers, music and visual media [4].

According to AP reporting, the department told the court that the “creative possibilities and public benefits” of training large language models on such content outweigh competitive harm to publishers, and warned that a ruling against OpenAI could hinder scientific progress, economic mobility and U.S. national-security interests [2]. Reuters similarly reported that the filing appears to be the first time the U.S. government has weighed in on the wave of copyright cases brought by authors, publishers, music labels and news outlets over AI training [5].

For the Seattle Times, that intervention raises the stakes. The newspaper is not only seeking damages or leverage in a licensing dispute. It is asking a court to recognize that journalism has a legally protected market value when it is copied into AI systems. The government is now urging a narrower view: training should be evaluated separately from later outputs that might reproduce or closely imitate protected expression [4].

That distinction could become decisive. If a court agrees that training alone is highly transformative fair use, publishers may have to focus more heavily on evidence of infringing outputs, removal of copyright-management information, substitution for news products, or unlawful acquisition of training data. If the court rejects that view, AI developers could face a much broader duty to license journalism before using it to train models.

The publisher’s argument: journalism is not free raw material

The Seattle Times’ position reflects a broader fear across the news industry: AI platforms may become answer engines that absorb reporting, summarize the world for users, and reduce the need to visit the publications that actually gathered the facts. That is especially sensitive for regional newsrooms, whose reporting on city halls, courts, schools, business and public safety is expensive to produce and difficult to replace.

The publisher’s claim is not simply that AI tools “learned” from its stories. It is that the learning process allegedly required copying copyrighted works, processing them at scale, and incorporating the value of those works into commercial systems sold or distributed by OpenAI and Microsoft. From the newspaper’s perspective, a subscription article, an investigation or a local business scoop is not equivalent to a loose fact in the public domain. It is a protected expression built through reporting, editing, verification and legal risk.

The New York Times, which is litigating closely watched claims against the same defendants, has framed the issue in similar terms. Its spokesperson said the administration was siding with “trillion-dollar AI companies” at the expense of creators whose work was allegedly taken, and argued that AI companies should pay fairly for the content that makes their products possible [2]. That statement is not from the Seattle Times, but it captures the publisher-side theory now being tested in the same legal ecosystem.

OpenAI and Microsoft’s likely defense

OpenAI and Microsoft have consistently defended AI training as a transformative use. Their core argument is that training a model is not the same as republishing an article. In that view, the model studies statistical relationships among words, facts and styles across a vast corpus, then uses those learned relationships to generate new text. TechCrunch reported that the government’s brief adopted that broad framing, saying the United States has a strong interest in a competitive AI industry and warning that constraining large-language-model development would undermine progress [3].

The companies are also likely to emphasize that copyright protects expression, not facts. News articles contain both. A report about a mayoral decision, a Boeing contract or a court ruling may include facts anyone can discuss, but the article’s structure, wording, selection and presentation are protected. The legal fight is therefore about whether model training copies protected expression in a way copyright law forbids, or whether the use is sufficiently transformative and non-substitutive to qualify as fair use.

Microsoft’s role is especially important because it is not only an investor in OpenAI. It has integrated OpenAI technology into products such as Copilot and has provided major computing infrastructure. Publisher lawsuits against both companies seek to connect Microsoft’s commercial deployment and technical support to the alleged copying and downstream market harm.

Why the Seattle Times case matters locally

The Seattle Times is not a distant observer of the AI economy. It covers Microsoft in its own region and reports on a technology ecosystem that includes cloud computing, software, startups and AI deployment. That makes the lawsuit symbolically sharp: a local newspaper is challenging one of the Pacific Northwest’s most powerful corporate institutions, along with the AI company whose tools have reshaped the information market.

The local dimension matters because regional journalism depends on a fragile economic model. National outlets may have large subscription bases, diversified products and global licensing opportunities. Local and regional papers often have fewer negotiating options. If AI companies can train on their archives without permission, those papers may lose bargaining power just as AI-generated answers begin to compete with search traffic and direct readership.

At the same time, the Seattle Times’ case lands in a complicated environment. Some publishers have signed AI licensing deals; others have sued. OpenAI has pursued partnerships with local news organizations while simultaneously facing litigation from publishers over alleged copyright infringement [6]. That split reveals an industry trying to decide whether AI companies are future distribution partners, existential competitors, or both.

The fair-use question courts must answer

Fair use is not a slogan. Courts usually weigh purpose and character of the use, the nature of the copyrighted work, the amount used, and the effect on the market for the original. In AI cases, those factors become difficult because training can require copying entire works, while the public output may not display the copied text.

The government’s filing urges courts to focus on transformation and to separate training from outputs [4]. Publishers will likely respond that this separation is artificial: without copying the articles, the model could not acquire the same ability to summarize, imitate or compete. They will also argue that licensing markets for news content already exist, and that AI companies should not be allowed to destroy or bypass those markets by calling their use transformative.

Recent AI copyright decisions have not produced a single clean rule. WIRED noted that judges have treated training questions differently depending on facts, including whether materials were lawfully acquired and whether plaintiffs showed market harm [4]. That uncertainty is why the Seattle Times case, and the related publisher actions, remain so important.

What happens next

The immediate takeaway is that the Seattle Times’ lawsuit has become harder to view in isolation. The Justice Department’s September 2026 filing gives OpenAI and Microsoft a federal policy endorsement for the proposition that AI training on copyrighted text can be fair use [5]. It also gives publishers a clearer target: they must persuade courts that journalism is not merely data, and that uncompensated training threatens a real market for licensed news content.

No court has yet delivered a final, industry-wide answer. Until one does, the Seattle Times case remains part of the central legal contest of the generative-AI era: whether the companies building the next layer of the internet must pay the institutions whose reporting helped make those systems useful.

Developments

1. Seattle Times and Newsday sue OpenAI and Microsoft over copyrightReuters · Sep 5, 2026, 3:01 AM UTC · 8/10
2. Seattle Times sues Microsoft and OpenAI over AI training on journalismGeekWire · Sep 5, 2026, 1:50 AM UTC · 9/10

Sources from the last 72 hours

1. [1]US government sides with OpenAI on issue of training LLMs on copyrighted materialSep 2, 2026, 5:09 PM UTC
2. [2]Trump Administration Sides With OpenAI in New York Times Copyright LawsuitSep 2, 2026, 6:41 PM UTC
3. [3]US government backs OpenAI in New York Times copyright caseSep 2, 2026, 2:53 PM UTC
4. [4]Seattle Times sues OpenAI, Microsoft over copyright infringementSep 4, 2026, 12:00 AM UTC
5. [5]Trump administration backs OpenAI in New York Times’ copyright case over training of chatbotsSep 2, 2026, 7:12 PM UTC

AI-generated article based on recent web research, then preserved as a dated editorial snapshot.