News Publishers Accuse AI Companies of Systematic Intellectual Property Theft
Major AI companies are using copyrighted journalism without permission or compensation to train their models, a practice that threatens the financial viability of news organizations worldwide, according to the publisher of The New York Times.
The companies-including OpenAI, Google, Meta, Microsoft, and Anthropic-strip-mine news websites to build their products, then present the repackaged content as their own. This happens not once during training, but continuously, every single day.
The damage extends beyond lost traffic. As AI chatbots answer questions directly instead of sending users to news sites, referral traffic from AI products runs 96 percent lower than Google search, according to industry research. The largest newspapers tracked by Comscore saw traffic drops exceeding 45 percent as the AI race intensified over the past four years.
The Economics Don't Add Up
The six leading AI companies have a combined valuation of $11 trillion. Private AI investment in the United States reached nearly $350 billion in 2025. Yet less than half of 1 percent of that investment appears to compensate the people and organizations creating the content that powers these systems.
AI executives have acknowledged that original, high-quality journalism is particularly valuable for model effectiveness. Five of the top 10 sites used to train popular large-language models belong to news publishers. OpenAI stated it would be "impossible to train today's leading AI models without using copyrighted materials."
The New York Times alone published nearly half a million original works last year-articles, photos, videos, and podcasts-at a cost exceeding $2 billion. The organization maintains journalists in all 50 U.S. states and 155 countries. Yet when The Times sued OpenAI and Microsoft for copyright violations, the litigation stretched over two and a half years and cost more than $20 million.
Why This Matters for Newsrooms
The U.S. news industry has already lost 75 percent of its journalists over the past two decades. More than 3,000 newspapers have closed. Another shuts down every three days.
Many news organizations turned to subscriptions to offset advertising declines-a decline of 80 percent combined over 20 years. But when people can access stolen work free through AI products, building subscriber relationships becomes harder. About 30 percent of AI bot scrapes violate explicit restrictions on accessing websites' content, including paywalled material.
Some larger news organizations have signed licensing deals with AI companies. But there's reason to question whether these payments will offset lost revenue and readers. The vast majority of news publishers say they expect no significant income from AI platforms.
The Quality Problem
Research by the European Broadcasting Union found that leading AI assistants significantly misrepresented news in nearly half of all answers. Google and Apple have made major errors using AI to rewrite news headlines and alerts.
AI systems tend to express uncertainty poorly, often delivering confident wrong answers. Unlike news organizations, AI companies don't track or correct errors, leaving users with no way to know when they've been misled.
A majority of online content is already generated by AI, according to Amazon Web Services estimates. Some experts expect that number to reach 90 percent in coming years. Meanwhile, fake local news sites now outnumber real ones as AI makes it cheaper to fabricate sites and harder for legitimate ones to survive.
What News Organizations Can Do
Stand up for intellectual property rights. These rights are anchored in law and ethics, but they only hold if enforced. This requires resources many organizations lack, but tolerating systematic theft eventually ends the ability to continue reporting.
Evaluate licensing deals carefully. Tech companies already possess the leverage of having taken your content. Before agreeing to terms, ask whether payment reflects fair value and whether you retain meaningful control over how your work is used.
Push legislators for clear protections. The AI industry spends heavily on lobbying and political donations. News organizations must work together to advocate for intellectual property protections, bot transparency requirements, and legal accountability for defamatory AI-generated content.
Use AI responsibly within your own newsroom. The technology itself isn't inherently harmful. Newsrooms should establish standards for responsible AI use to improve reporting, editing, and distribution while maintaining human decision-making.
Build direct audience relationships. A world mediated by AI platforms leaves news organizations at the mercy of tech giants for traffic and credit. The clearest path forward is making your journalism so distinctive it has its own gravity-forcing audiences to seek you directly rather than through intermediaries.
Focus on original reporting. Aggregation and clickbait worked in the search and social media era. They won't survive AI competition. AI systems cannot do original reporting. They mine the public record but cannot add to it. That's where journalism's irreplaceable value lies.
The Broader Stakes
When local news organizations fail, communities see declining civic engagement, rising public corruption, and eroding trust. Evidence shows people trust each other less and hate each other more when local reporting disappears.
The public understands the stakes. Two-thirds of Americans are highly concerned about AI spreading inaccurate information, according to Pew Research. Yet when people want to fact-check something, their preferred option is "a news source I trust"-not an AI chatbot.
The tech giants argue they have no obligation to the creators whose work powers their products. They simultaneously market their commitment to journalism while fighting in court to avoid paying for it.
The question before news organizations is whether the value created by quality reporting will be captured by tech giants or returned to newsrooms to sustain essential work. The answer will determine whether original journalism survives the AI era.
For writers specifically, understanding how AI for Writers intersects with these business and legal questions is essential. The broader context of generative AI and LLM systems helps explain why these disputes matter for the future of professional writing.
Your membership also unlocks:
