Seattle Times, Newsday Sue OpenAI & Microsoft Over AI Training

In a significant escalation of the ongoing legal battles surrounding artificial intelligence, The Seattle Times and Newsday have filed a joint lawsuit against OpenAI and Microsoft. The two prominent news organizations allege the tech giants systematically scraped their journalistic content, bypassed paywalls, and used this material to train their AI models, thereby creating AI-generated alternatives that cannibalize traffic and advertising revenue.

The complaint, filed on September 5, 2026, claims OpenAI and Microsoft's actions constitute copyright infringement. It highlights that the method of acquiring content involved circumventing access restrictions, a practice explicitly discouraged by international AI codes. The newsrooms critically labeled generative AI "a snake eating its own tail," pointing to the models' tendency to also "hallucinate," attributing false information back to the original publishers while stripping away crucial copyright management data.

This lawsuit adds considerable weight to the growing chorus of media outlets challenging the training practices of leading AI developers. It follows a similar copyright infringement suit filed by The New York Times in 2023, and aligns with legal action already undertaken by over 400 local newspapers. In contrast, other major news organizations, such as the Associated Press and Vox Media, have opted to license their content archives directly to OpenAI, indicating a bifurcated response from the publishing industry.

A key allegation in the Seattle Times and Newsday complaint centers on the circumvention of paywalls and subscription models. This claim goes beyond merely what was done with the content, focusing on how the material was obtained. This aspect carries particular significance, especially in Europe, where OpenAI is a signatory to the EU's general-purpose AI code. This code, effective since August 2025, specifically commits signatories not to bypass effective technological measures that restrict access, with paywalls and subscription models explicitly mentioned.

The EU code further mandates that crawlers must adhere to robots.txt protocols and avoid sites identified by courts or public authorities as persistent infringers on a commercial scale. While the code doesn't retroactively apply to conduct predating August 2025, and the legal recognition of a paywall as a "reservation of rights" under the 2001 copyright directive remains unsettled, OpenAI's adherence to these principles is now under scrutiny.

Simultaneously, Microsoft, a co-defendant, sought to dismiss two core theories underpinning such complaints—lost licensing fees and market dilution—in the same court on the same day. However, their defense brief, covered by TNW earlier, predominantly focused on books rather than the specific means of accessing web pages, leaving the paywall circumvention argument largely unaddressed by their current public position. The legal landscape in the U.S. differs significantly from Europe, where the concept of "fair use" has no direct equivalent. A July ruling by a German court in a case brought by GEMA already established that works memorized within model parameters infringe copyright at the point of storage, setting a precedent that European publishers could leverage.

The increasing number of lawsuits underscores the media industry's demand for fair compensation and protection against what they view as unauthorized use of their intellectual property. The outcomes of these cases could fundamentally reshape how AI models are trained, potentially forcing developers to secure licenses for copyrighted material more broadly and implement stricter protocols to respect digital access restrictions. The legal precedents set in both the U.S. and Europe will be critical in defining the future economic relationship between content creators and generative AI developers.

FAQ

Q: What are the main allegations against OpenAI and Microsoft?

A: The lawsuit alleges systematic scraping of news articles, circumvention of paywalls, using this content to train AI models, creating AI-generated alternatives that reduce traffic and revenue for newsrooms, attributing false information (hallucinations), and stripping copyright management information from articles.

Q: How does this lawsuit relate to previous legal actions against AI companies?

A: The Seattle Times and Newsday join The New York Times, which sued in 2023, and 400 local newspapers already involved in similar litigation. Conversely, some organizations like AP and Vox Media have chosen to license their content to OpenAI.

Q: What is the significance of the paywall circumvention claim, especially concerning European regulations?

A: The paywall circumvention claim is significant because it addresses how the training data was obtained, not just its use. OpenAI is a signatory to the EU's general-purpose AI code, which commits companies not to bypass access restrictions like paywalls, a measure that has been in effect since August 2025.