OpenAI and Microsoft Face Copyright Lawsuit from Seattle Times

TL;DR: The Seattle Times and Newsday sued OpenAI and Microsoft over unauthorized data scraping to train generative AI models. This lawsuit signals a critical expansion of media resistance from national giants to regional publishers, potentially redefining fair use precedents for the entire journalism industry.

Key facts

• The Seattle Times and Newsday filed a federal copyright lawsuit against OpenAI and Microsoft on September 4, 2026, in the U.S. District Court for the Southern District of New York.
• Plaintiffs allege that tech giants scraped journalism from their websites without permission or payment to train generative AI models, including ChatGPT, GitHub Copilot, and Bing’s AI features.
• The lawsuit marks a shift in media resistance, expanding beyond national outlets like The New York Times to include regional publishers reliant on subscription revenue.
• The complaint argues that AI systems function as “rapacious consumers” by ingesting human-authored work behind paywalls and returning derivative imitations for commercial gain.
• Plaintiffs are seeking a court order to destroy copies of their works and any datasets or AI models containing this material, likening the process to a “snake eating its own tail.”
• Microsoft and OpenAI have previously funded journalism fellowships at The Seattle Times, creating tension between corporate partnerships and publishers’ rights to control content.
• A Microsoft spokesperson expressed surprise at the lawsuit but remains open to discussing solutions, while OpenAI maintains its models are trained on publicly available data under fair use.

Media Outlets Expand Copyright Lawsuit Against AI Giants

The Seattle Times and Newsday have joined a growing legal coalition against OpenAI and Microsoft, filing a federal copyright lawsuit in the U.S. District Court for the Southern District of New York on September 4, 2026 [8]. The complaint alleges that both tech giants scraped journalism from their websites without permission or payment to train generative AI models, including ChatGPT, GitHub Copilot, and Bing’s AI features [3][5][7].

This litigation marks a significant shift in the copyright debate. While previous lawsuits were driven by major national outlets like The New York Times, the inclusion of regional publishers signals that the resistance to unlicensed AI training is spreading beyond elite media institutions to local newsrooms that rely heavily on subscription revenue [1][4][6].

Allegations of Unauthorized Data Scraping

The core of the lawsuit centers on how OpenAI and Microsoft acquired content. The plaintiffs argue that the companies harvested articles, including those behind paywalls, to build their training datasets [8]. They contend that these AI products are marketed as tools for creating new content but function instead as “rapacious consumers” [1][4][6].

According to the complaint, these systems ingest human-authored work and return derivative imitations for commercial gain. The newspapers assert that this practice reduces user traffic and subscription rates by allowing users to access summaries or paraphrased versions of their reporting without paying [8].

The lawsuit describes generative AI as a self-consuming cycle. It argues that the technology threatens to fracture the traditional news ecosystem by destroying the original reporting it relies upon, likening the process to a “snake eating its own tail” [1][4][6]. The plaintiffs are seeking a court order to destroy copies of their works and any datasets or AI models containing this material [8].

Internal Costs and Financial Tensions

The financial stakes for traditional media are high. Alan Fisco, CEO of The Seattle Times, stated in an internal memo that the publication spends millions annually on content creation and must protect it from unauthorized use [8]. This economic pressure is a key driver behind the legal action.

The case carries particular complexity due to the existing financial relationship between the publisher and the defendants. Microsoft and OpenAI have previously funded journalism fellowships and newsroom innovation initiatives at The Seattle Times [1][3][4][5][6]. This prior collaboration has drawn attention from industry observers, highlighting the tension between corporate partnerships in media innovation and the fundamental rights of publishers to control their content.

Precedent and Response

This lawsuit follows a landmark 2023 copyright infringement suit filed by The New York Times against OpenAI and Microsoft [1][3][4]. The Seattle Times case adds weight to the argument that AI developers are not just using public data, but specifically targeting proprietary journalistic content.

In response to the filing, a Microsoft spokesperson told GeekWire that the company was “surprised by the lawsuit” but remains open to discussing solutions [1][4][6]. An OpenAI representative stated that their models are trained on publicly available data under the fair use doctrine and did not comment specifically on the suit [8].

Legal experts note that this case could establish critical precedents regarding fair use, data scraping rights, and the economic sustainability of professional journalism in an automated landscape [3]. As more regional publishers consider similar actions, the outcome may define how AI companies can legally access high-quality human content for training purposes.

Sources

8. Seattle Times and Newsday sue OpenAI and Microsoft in the US (ua.news) — 2026-09-05
9. HyperAI (hyper.ai) — 2026-09-06
10. Seattle Times and Newsday are the latest publications to sue OpenAI and Microsoft | daily.dev — 2026-09-05
11. Seattle Times, Newsday Sue OpenAI, Microsoft Over AI Training (www.techbuzz.ai) — 2026-09-05
12. Seattle Times and Newsday are the latest publications to sue OpenAI and Microsoft | TechCrunch (techcrunch.com) — 2026-09-05
13. Seattle Times and Newsday are the latest publications to sue OpenAI and Microsoft (www.yahoo.com) — 2026-09-05
14. Seattle Times and Newsday are the latest publications to sue OpenAI and Microsoft | TechCrunch (techcrunch.com) — 2026-09-05