Across the digital commons, a reckoning is taking shape: major news organizations, including The New York Times and The Economist, have united in 2026 to demand that artificial intelligence companies cease training their systems on published journalism without consent or compensation. The coalition represents a maturation of the publishing industry's response to a practice that has quietly extracted the fruits of human inquiry and transformed them into commercial AI products. At its core, this is an ancient question wearing new clothes — who owns the labor of the mind, and who may profit from
Major Publishers Form Coalition to Demand Compensation for AI Training Data
Publishers are asking: why should our work be free?
Mark
So these publishers are saying AI companies stole from them. What exactly did the AI companies do?
Mimi
They took published articles—journalism that reporters wrote and editors refined—and fed them into the systems that power AI models. The AI learned patterns from that text without asking permission or paying anything.
Luke
That's the claim, but we should be precise about what "using" means here. Did the AI companies directly copy articles into their products, or did they use them as training data that the model learned from?
Mimi
Training data. The model learned from the text, but the articles themselves aren't stored or reproduced in the final product.
Luke
Right. Which is why the legal question is actually hard. Fair use doctrine might protect that kind of learning. We don't know yet.
Mark
Why does it matter that these organizations joined together instead of suing separately?
Mimi
Leverage. One publisher suing is expensive and risky. A coalition of major news organizations—the Times, the Economist, others with resources and reach—that's a different negotiating position.
Luke
Though we should note: the article doesn't specify which other publishers are in the coalition or how many. It names two. That's worth flagging.
Mark
What do the publishers actually want?
Mimi
Compensation and permission. They want AI companies to license their content and pay for it, the way you'd pay for any other intellectual property.
Luke
And if the AI companies refuse?
Mimi
Then presumably litigation. But again, the legal outcome is genuinely uncertain.
Mark
What happens if the publishers win?
Mimi
It could reshape how AI companies build their systems. They'd have to negotiate with creators instead of just scraping the internet for free.
Der Puls
- Major publishers have stopped absorbing the loss quietly — they are now organizing collectively, treating unauthorized AI data scraping as an existential economic threat rather than a technical nuisance.
- The tension cuts deep: AI systems trained on years of investigative journalism can now answer questions and summarize findings without ever sending a reader to a paywall, quietly eroding the subscription models that fund the work itself.
- The coalition's power lies in its coordination — where a single publisher faces ruinous legal costs and uncertain outcomes, a unified bloc of globally recognized outlets forces tech companies to recalculate the cost of resistance.
- The legal ground remains genuinely contested, with AI companies betting on fair use protections while some have already begun negotiating licenses — a signal that the threat is being taken seriously behind closed doors.
- The outcome will set precedent far beyond journalism, touching every creator whose work has been absorbed into AI training datasets, and defining whether the digital age treats human creativity as a commons or a contract.
Across the digital commons, a reckoning is taking shape: major news organizations, including The New York Times and The Economist, have united in 2026 to demand that artificial intelligence companies cease training their systems on published journalism without consent or compensation. The coalition represents a maturation of the publishing industry's response to a practice that has quietly extracted the fruits of human inquiry and transformed them into commercial AI products. At its core, this is an ancient question wearing new clothes — who owns the labor of the mind, and who may profit from it?
The New York Times, The Economist, and a growing number of major news organizations have formed a coalition to confront artificial intelligence companies that train their systems on published journalism without permission or payment. It is among the most organized responses yet from the content industry to the tech sector's data practices — a declaration that publishers no longer intend to serve as free fuel for AI development.
For years, news organizations watched as their reporting was scraped from the internet and fed into large language models. These systems learn from vast quantities of text, much of it journalism that reporters spent months producing and editors hours refining. Publishers call it intellectual property theft. What distinguishes this coalition from earlier individual complaints is its scale: together, these organizations possess the legal resources, public platforms, and collective leverage that no single outlet could muster alone. The message is direct — if you want to train on our work, you will negotiate, and you will pay.
For outlets like The Times and The Economist, the stakes are not merely philosophical. Both have built their futures on digital subscriptions and premium content. An AI that can summarize their investigations without routing readers back to their paywalls threatens the financial model that makes serious journalism possible in the first place.
The coalition also surfaces a broader question about the digital economy: tech companies have built trillion-dollar valuations by aggregating information created by others. Publishers are now asking why their work should be treated as a free commons while AI companies charge for access to systems trained on that very work.
The legal outcome remains uncertain. Tech companies argue that training on publicly available text constitutes fair use, and courts have not yet settled the question definitively. Some AI developers have begun licensing content, suggesting they regard the legal risk as real. Others are holding firm. The coalition's next moves will involve both litigation and negotiation — and how the industry responds will define not just the future of news, but the relationship between human creativity and the companies building the next generation of intelligence.
The New York Times, The Economist, and a growing roster of major news organizations have banded together to confront what they see as a fundamental threat to their business: artificial intelligence companies training their systems on published journalism without permission or payment. The coalition represents one of the most organized pushbacks yet from content creators against the tech industry's data practices, signaling that publishers are no longer willing to treat their work as free fuel for AI development.
The formation of this group marks a shift in how the publishing industry is responding to AI. For years, news organizations have watched as their articles, investigations, and reporting have been scraped from the internet and fed into large language models—the underlying technology behind systems like ChatGPT and Claude. These models learn patterns from vast amounts of text, and much of that text comes from journalism that reporters spent months researching and editors spent hours refining. The publishers argue that this constitutes theft of intellectual property, a taking of their most valuable asset without negotiation or recompense.
What distinguishes this coalition from earlier individual complaints is its scale and coordination. By joining forces, these organizations are attempting to create leverage they lack alone. A single publisher suing an AI company faces enormous legal costs and uncertain outcomes. A coalition of major news outlets—organizations with global reach, legal resources, and public platforms—presents a different calculus for tech companies considering their licensing strategies. The group is essentially saying: if you want to train your AI on our work, you will need to negotiate with us, and you will need to pay.
The Economist and the New York Times are particularly significant members of this coalition. Both organizations have invested heavily in digital subscription models and premium content strategies. For them, the unauthorized use of their journalism in AI training represents not just a philosophical violation but a direct economic threat. If an AI system can summarize their reporting or answer questions based on their investigations without directing readers back to their paywalls, the incentive for people to subscribe diminishes. The financial model that sustains quality journalism—the one that pays reporters to spend weeks on investigations—depends on readers valuing that work enough to pay for it.
The coalition's formation also reflects a broader reckoning about who owns and profits from data in the digital age. Tech companies have built trillion-dollar valuations partly by aggregating and monetizing information created by others. Publishers are now asking: why should we be the exception? Why should our work be treated as a commons available for free extraction while AI companies charge for access to systems trained on that very work? The question extends beyond journalism to authors, artists, and other creators whose work has been similarly incorporated into AI training datasets without consent or compensation.
What remains unclear is how much leverage the coalition actually possesses. Tech companies have argued that training on publicly available text falls within fair use—a legal doctrine that permits certain uses of copyrighted material without permission. Courts have not yet definitively ruled on whether AI training constitutes fair use, and the legal landscape remains unsettled. Some AI companies have begun negotiating licensing agreements with publishers, suggesting they view legal risk as real. Others have resisted, betting that fair use will ultimately protect them.
The coalition's next moves will likely involve both legal action and negotiation. Publishers are signaling that they expect AI companies to either license their content or stop using it. How the tech industry responds—whether through licensing agreements, technical measures to exclude copyrighted material, or legal challenges to the publishers' claims—will shape not just the future of news but the broader relationship between creators and the companies building AI systems. For now, the coalition has drawn a line. Whether it holds depends on what happens next.
Bemerkenswerte Zitate
Publishers are signaling that AI companies must either license their content or stop using it— Coalition position
Möchten Sie die ganze Geschichte?
Das Original lesen bei Reuters ↗
Technology