Google pays 400 publishers to feed its AI news features
Google is paying over 400 news outlets to power AI summaries and audio briefings, keeping readers on its own platform.
Business · Source: Google (The Keyword)
What happened
Google just expanded its News AI pilot program to include more than 400 publications worldwide. The tech giant is officially paying news outlets to feed their content into experimental AI features. New partners include global heavyweights like CNN, Axios, Forbes, and The Financial Times. Local newsrooms are also signing deals, including the Texas Tribune, Oil City News in Wyoming, and The Post and Courier in South Carolina.
The pilot program originally launched in December 2025 with a smaller group of publishers like The Washington Post, The Guardian, and El Pais. Google uses the licensed content to build and test new AI tools directly inside the Google News ecosystem. One feature generates in-depth article summaries on publisher pages to give readers context before they click. Another feature creates customizable audio briefings, acting like an AI podcast that stitches together daily headlines based on user interests.
Google refuses to disclose the exact financial terms of these licensing deals. However, a company spokesperson confirmed to Nieman Lab that the program does not use a pay-per-value model. Publishers are not paid based on how often their specific articles are used to generate AI outputs. Instead, they receive a flat payment to provide enhanced content for Google to train and run its experimental features.
Key facts
- 400 — Number of publications now participating in the Google News AI pilot program.
- December 2025 — When Google originally launched the News AI pilot program.
Why it matters
The era of free web scraping is closing. Massive tech companies are realizing that high-quality, real-time data requires formal licensing agreements. Google is setting a clear precedent that if you want premium news content to power your AI, you have to pay the creators. This shifts the foundation of how AI models consume the internet. It also creates a massive financial moat. Google can easily afford to cut checks to 400 different publishers. Independent builders and early-stage startups cannot.
For the publishers, this deal is a double-edged sword. They finally get paid for their content after years of watching AI tools scrape their websites for free. But the payment does not shift the actual balance of power. Google uses the licensed content to build AI summaries and audio briefings that keep users inside the Google News app. The AI gives the reader the whole story upfront. The reader never needs to click the link. Publishers get a short-term check, but they sacrifice their direct relationship with the audience.
For builders
Licensing costs create a massive startup moat
Google is normalizing paid data deals for AI training and output. Small startups building news or research agents will struggle to afford these licenses. The big tech companies win by pricing indie developers out of the premium data market.
Audio generation is the new text summary
Google is heavily pushing customizable AI podcasts that read the news to users. Text summaries are no longer enough to capture user attention. Builders need to integrate audio generation into their products to match shifting consumer habits.
Flat fees beat pay-per-value models
Google intentionally avoided a pay-per-value structure for this pilot. Tracking exactly whose data generated an AI response is too complex. Builders licensing data should negotiate flat access fees to avoid massive engineering overhead.
My take
Google is buying off the exact publishers its AI is designed to replace. The newsrooms get a quick check, but Google gets to keep the reader trapped on its own platform forever. It is a brilliant business move by Google and a desperate surrender by the media industry.
Original reporting: Google (The Keyword). This is my rewrite and opinion.