In a moment that may define the relationship between journalism and artificial intelligence for years to come, two major American regional newspapers — the Seattle Times and Newsday — have filed separate lawsuits against OpenAI and Microsoft, alleging that paywalled articles were harvested without consent to train AI systems. The cases arrive as news organizations grapple with a quiet erosion of their economic foundations, watching readers migrate toward AI-generated summaries rather than the original reporting those summaries depend upon. At stake is not merely a legal dispute over licensing,
Seattle Times, Newsday Sue OpenAI and Microsoft Over Paywalled Content
Journalism behind a paywall has a different legal status than freely published material.
Mark
What exactly are the newspapers claiming happened here? Did OpenAI and Microsoft just copy their articles?
Mimi
They're saying the companies scraped their paywalled articles—the ones behind subscription walls—and used them to train AI models without permission or payment. It's not that the articles appear in ChatGPT's output; it's that they were part of the training data.
Luke
Right, and that's important to be clear about. The lawsuits don't allege that OpenAI is republishing the articles. They're alleging the articles were used as input to build the models. Those are different claims legally.
Mark
Why does it matter that the articles were paywalled?
Mimi
Because paywalled content is how these newspapers generate revenue. Readers pay to access it. If AI companies can use that content for free to train systems, it undermines the business model that funds their newsrooms.
Luke
That's the newspapers' argument, and it's economically sound. But legally, it's not clear that paywalled status changes the copyright analysis. Fair use doctrine doesn't typically distinguish between free and paid content.
Mark
What's the fair use argument the AI companies would make?
Mimi
They'd say using published material to train AI is transformative—it's not republishing the articles, it's using them as raw material for a new technology. Fair use permits that kind of copying for certain purposes.
Luke
And they have case law on their side. Courts have been fairly permissive about fair use in technology contexts. But this is untested territory. No court has ruled on whether training generative AI qualifies as fair use.
Mark
If the newspapers win, what changes?
Mimi
AI companies would likely need to license content from publishers before using it in training. That could mean paying news organizations, or at least getting permission. It would reshape how these companies build their models.
Luke
Or it could mean they train on different data—older material in the public domain, or content from sources willing to license cheaply. The outcome isn't predetermined.
Mark
How long will this take?
Mimi
These cases typically take years. We're probably looking at several years of litigation, possibly appeals. But the precedent could be significant for the entire industry.
Luke
And that's worth noting: the outcome here will likely influence how other publishers approach this. If the newspapers win, you'll probably see a wave of similar suits. If they lose, publishers may shift strategy entirely.
The Pulse
- Two of America's largest regional newspapers have moved from grievance to litigation, accusing OpenAI and Microsoft of systematically mining subscription-only journalism to fuel AI model training without payment or permission.
- The targeting of paywalled content sharpens the conflict considerably — these are not freely published articles but journalism that readers and institutions pay to access, representing years of deliberate investment in subscription infrastructure.
- Publishers are watching their web traffic hollow out as AI chatbots absorb and redistribute their reporting, and some have concluded that licensing negotiations alone cannot arrest the damage.
- OpenAI and Microsoft are expected to invoke fair use doctrine as their primary defense, setting up a judicial test that could either constrain or legitimize the industry's current approach to training data sourcing.
- With the New York Times already in similar litigation, these cases are converging toward a potential legal reckoning that could force AI companies to negotiate licenses — or confirm they need not.
In a moment that may define the relationship between journalism and artificial intelligence for years to come, two major American regional newspapers — the Seattle Times and Newsday — have filed separate lawsuits against OpenAI and Microsoft, alleging that paywalled articles were harvested without consent to train AI systems. The cases arrive as news organizations grapple with a quiet erosion of their economic foundations, watching readers migrate toward AI-generated summaries rather than the original reporting those summaries depend upon. At stake is not merely a legal dispute over licensing, but a deeper question about who bears the cost of knowledge — and who has the right to profit from it.
The Seattle Times and Newsday have each filed lawsuits against OpenAI and Microsoft, alleging that the companies systematically used their paywalled articles — without permission or compensation — to train generative AI systems. Together serving millions of readers across Washington State and New York, the two newspapers are seeking damages for what they describe as large-scale copyright infringement at the core of how AI models are built.
What distinguishes these cases from earlier content-scraping complaints is their focus on paywalled material specifically. Newspapers have spent years constructing subscription models to replace lost advertising revenue, and the lawsuits argue that incorporating that protected journalism into AI training datasets allowed OpenAI and Microsoft to profit from work they never paid for. The companies are expected to argue fair use — the legal doctrine permitting limited reproduction for purposes like commentary or education — but whether courts will extend that doctrine to AI training remains genuinely unsettled.
The broader context is difficult to ignore. Publishers have watched readers increasingly turn to AI chatbots for information rather than visiting news sites directly, compressing both traffic and subscription incentives. For organizations like the Seattle Times and Newsday, which depend on subscription revenue to fund original local reporting, the stakes are existential as much as legal.
The New York Times has pursued a parallel legal strategy, and the cumulative weight of these cases could reshape how AI companies source training data industry-wide. A ruling in favor of the newspapers might require AI developers to license copyrighted material before use; a ruling upholding fair use would largely preserve the status quo. Either outcome will take years to arrive — but the filing of these suits signals that major news organizations have decided waiting is no longer an option.
Two of the country's largest regional newspapers have filed separate lawsuits against OpenAI and Microsoft, accusing the companies of systematically harvesting their paywalled articles to train artificial intelligence systems without permission or compensation. The Seattle Times and Newsday, which together serve millions of readers across Washington State and New York, are seeking damages for what they characterize as large-scale copyright infringement—a legal challenge that strikes at the heart of how generative AI companies source the text they use to build their models.
The lawsuits represent a significant escalation in the conflict between news organizations and AI developers over intellectual property rights. Unlike earlier complaints that focused on general concerns about content scraping, these cases target a specific practice: the unauthorized use of articles behind paywalls, the digital barriers that newspapers have spent years building to generate subscription revenue. The companies argue that their paywalled content—journalism that readers pay to access—was incorporated into training datasets without consent, effectively allowing OpenAI and Microsoft to profit from work that the newspapers invested substantial resources to produce.
This legal action arrives as the news industry faces mounting pressure from AI systems that can summarize, rewrite, and synthesize reporting at scale. Publishers have watched their traffic patterns shift as readers increasingly turn to AI chatbots for information rather than visiting news websites directly. The lawsuits suggest that some of the largest regional news organizations have concluded that negotiation and licensing agreements are insufficient—that only litigation can establish whether AI companies have a legal obligation to compensate them for the content used in training.
The cases invoke copyright law as the primary legal framework, but the filings also touch on broader questions about culture, sports coverage, and the role of local journalism in the digital age. The Seattle Times and Newsday both produce original reporting that serves their communities—investigations, breaking news, analysis—and the lawsuits assert that this work has distinct value that should not be appropriated without permission. The companies being sued have previously argued that their use of published material falls within fair use doctrine, a legal principle that permits limited copying for purposes like criticism, commentary, and education. Whether courts will agree remains an open question.
The timing of these lawsuits reflects a moment of reckoning in the AI industry. Other news organizations, including The New York Times, have pursued similar legal strategies, and the outcomes of these cases could reshape how AI companies source training data and whether they must negotiate licenses with publishers. If the courts rule in favor of the newspapers, it could establish a precedent requiring AI developers to obtain permission and potentially pay for copyrighted material used in model training. Conversely, if the courts uphold fair use defenses, it would signal that AI companies can continue their current practices with limited legal constraint.
For the Seattle Times and Newsday, the lawsuits represent a bet that the courts will recognize paywalled content as a distinct category deserving protection—that journalism behind a paywall has a different legal status than freely published material. Both newspapers depend on subscription revenue to fund their newsrooms, and they argue that allowing AI companies to use that paywalled content without compensation undermines their business model and, by extension, their ability to invest in journalism. The cases will likely take years to resolve, but they signal that major news organizations are no longer willing to accept the status quo.
Notable Quotes
The newspapers argue that allowing AI companies to use paywalled content without compensation undermines their business model and their ability to invest in journalism.— Seattle Times and Newsday position in lawsuits
Want the full story?
Read the original at Google News ↗
Technology