Study Finds AI Content Farms Now Flood Google News, Collect Ad Revenue From AT&T, Expedia, YouTube

NewsGuard and AI detector Pangram Labs have launched a real-time system to identify websites mass-producing AI-generated content, and the scale of what they found is hard to ignore: 3,006 “AI content farms” already flagged, a count that more than doubled in a single year and is currently growing by 300 to 500 new sites per month.

The system pairs NewsGuard’s media analysis data with Pangram Labs’ automated detection software. Pangram’s AI scans sites for content that appears machine-generated; human analysts then review flagged results to filter out false positives. A site gets labeled an AI content farm when it meets three conditions: a significant share of its content is AI-generated, the site doesn’t disclose that fact, and the layout is designed to look like human journalism.

The generic names — “Times Business News,” “Business Post” — are a tell, as are publishing schedules that push out dozens of articles per day.

Fake Coca-Cola Threats, Fabricated Senate Spending, and Russian Amplification

In October 2025, a farm called “News 24” published a false report claiming Coca-Cola threatened to pull its Super Bowl sponsorship if Puerto Rican rapper Bad Bunny performed at halftime. Coca-Cola doesn’t even sponsor the Super Bowl. Despite that, ads from Expedia, AT&T, YouTube, Priceline, Hotels.com, Skechers, and GoDaddy all ran on the site. Another farm, “CitizenWatchReport,” falsely claimed that two U.S. senators spent $814,000 on hotels in Ukraine. Russian state media picked up the story and pushed it back into American news feeds.

Of the identified farms, 358 were tied to Storm-1516, a pro-Russian influence operation that builds sites designed to look like local newspapers in the U.S. and Europe. NewsGuard notes the real number is likely higher, since detection methods are not foolproof. Iran and China-linked operations also show up in the data.

The advertising pipeline is what keeps these operations running. Most AI content farms are built specifically as “made for advertising” pages — cheap infrastructure designed to capture programmatic ad spending from brands that have no idea where their money is going.

Google’s own ad network, through AdSense, appears to profit from some of these sites, and AI content farms surface regularly in Google News and Google Discover without being filtered out. NewsGuard’s new data feed is designed to plug into ad-buying platforms like The Trade Desk, giving brands a way to block their placements from appearing on flagged sites.

The web has dealt with spam and SEO manipulation for decades. What AI changes is the unit economics: generating a fake news article now costs fractions of a cent and takes seconds. At 300 to 500 new farms per month, detection tools are in a race where the content side of the equation has a structural cost advantage that isn’t going away.