Google just won the legal right to train on your content without paying you, then convinced a hundred publishers to accept payment anyway.
The Summary
- Judge Amit Mehta dismissed antitrust suits from Chegg and Penske Media, ruling their claims that Google coerced free content for AI Overviews don't meet antitrust standards
- Google quietly launched a pilot paying ~100 publishers for content used in AI Overviews, AI Mode, and Gemini — one publisher earned over $1 million in the program's first year
- The timing is perfect: Google loses nothing legally while buying goodwill from publishers who now know lawsuits won't work
The Signal
US District Judge Amit Mehta ruled Wednesday that allegations of Google abusing monopoly power to extract free training data don't hold up under current antitrust law. Chegg and Penske Media Corporation argued Google forced publishers into an impossible choice: give us your content for AI training or vanish from search results. The judge disagreed.
The lawsuits, filed last year, centered on a core tension in the agent economy. When your AI assistant summarizes a Rolling Stone article or paraphrases Chegg's study guides directly in search results, users never click through. Publishers lose traffic, ad revenue, and subscription conversions. Google gets free training data and keeps users inside its walled garden.
"The judge sided with Google: coercing free content for AI features doesn't meet the legal bar for antitrust abuse."
But here's where it gets interesting. According to The Information, Google started a pilot program less than a year ago, before the lawsuit dismissal, paying roughly 100 publishers for contributions to AI Overviews, AI Mode in Search, and Gemini. One participating publisher pulled in over $1 million. The payments track how much each publisher's content contributed to AI-generated responses.
Key details:
- Pilot launched under 12 months ago, while lawsuits were active
- Covers three products: AI Overviews, AI Mode in Search, Gemini chatbot
- Payment tied to contribution volume, not clicks or impressions
- Digiday broke the story first, The Information confirmed scale
Google doesn't need to pay. The judge just said so. But they're paying anyway because the optics of a legal win while simultaneously compensating publishers sends a clear message: we'll pay when we want to, not when courts force us to. This isn't OpenAI scraping the web and waiting for lawsuits. This is platform power used surgically.
The distinction between these approaches matters. OpenAI fights publishers in court and loses licensing deals. Google beats publishers in court, then selectively pays some of them enough to make the others jealous. The pilot creates a class system: publishers inside the program versus publishers outside it, all watching their traffic decline while Google decides who gets compensated.
The Implication
If you're a publisher, the lesson is brutal: antitrust law won't save you from AI summarization eating your traffic. Your options are negotiate for a spot in Google's payment program, build direct audience relationships that bypass search entirely, or watch your business model dissolve.
For everyone else, this is the template for how platform power works in Web4. The agents scraping your data don't need your permission. The law won't make them pay. If they offer you money, it's strategy, not obligation. Own your distribution or own nothing.