NewsGuard Launches AI Detection Tool to Combat Misinformation Websites

NewsGuard has unveiled a new AI content farm detection tool aimed at stemming the rising tide of artificially generated misinformation flooding the digital news ecosystem.

Launched Thursday in collaboration with AI detection startup Pangram Labs, the tool identifies websites that primarily publish content created by large language models like ChatGPT, Claude, or Gemini. The system represents a technological countermeasure to an increasingly sophisticated problem that threatens to undermine information integrity online.

“If we can’t detect AI content, then every communication space is going to be flooded with inauthentic content that’s cheap to produce and difficult to impossible to differentiate from something authentic,” said Max Spero, Pangram’s CEO.

The detection process begins with Pangram’s proprietary AI models scanning entire domains for machine-generated content. Sites flagged by the algorithm undergo manual review by NewsGuard analysts, who verify the prevalence of AI content, check for transparency disclosures, and contact site owners for comment to prevent false positives.

NewsGuard categorizes websites as AI content farms based on three criteria: a substantial portion of AI-generated content as determined by Pangram’s technology; lack of disclosure about AI authorship; and presentation that could mislead average users into believing the content was human-created.

The system, which has undergone more than six months of testing, has already identified approximately 3,000 AI content farms—more than double what NewsGuard previously found using primarily manual techniques. Many of these sites operate under generic news-sounding names like “Times Business News” or “Business Post” while spreading misinformation about brands, political figures, and public health matters.

In one documented case, a site called Citizen Watch Report falsely claimed that Senators Lindsey Graham (R-SC) and Richard Blumenthal (D-CT) spent $814,000 on hotels in Ukraine. The fabricated story gained traction on social media and was amplified by Russian state media before being debunked.

Another example involved a site called News 24 falsely reporting that Coca-Cola threatened to withdraw its Super Bowl sponsorship over Bad Bunny’s halftime show performance—despite Coca-Cola not being a Super Bowl sponsor at all. The page displayed advertisements from major brands including AT&T, YouTube, Expedia, and Skechers.

Many of these websites fall into the category of “made-for-advertising” (MFA) sites—low-quality content operations designed solely to generate advertising revenue through traffic arbitrage.

“It’s a way to produce low-quality content for really low cost and generate some advertising revenue,” explained Matt Skibinski, NewsGuard’s chief operating officer. “Also, bad actors who want to spread false information have figured out that they can weaponize this technology and churn out, at a really high volume, false and misleading content and still make a quick buck because they run ads on those pages, too.”

The scale of the problem has grown dramatically in recent years. According to Pangram, between 300 and 500 new AI content farms emerge monthly. During a two-month observation period, NewsGuard identified 141 blue-chip brands inadvertently advertising on these sites.

NewsGuard hopes its detection tool will help both advertisers and consumers avoid AI-generated misinformation. The company will allow advertisers to license its data stream directly or through their agencies and has integrated with popular demand-side platform The Trade Desk to enable pre-bid blocking of these sites. The company is also considering adding the tool to its browser extension to help everyday users identify AI-generated content.

Pangram Labs, founded in 2023 by former engineers from Google and Tesla, has already received recognition for its technology’s effectiveness. A report in Nature highlighted Pangram’s capability to identify LLM-generated text in research papers and peer reviews. Several academic institutions have adopted the technology to combat undisclosed AI content in academic settings.

As AI-generated content proliferates, Spero anticipates growing demand for detection solutions: “There’s just going to be so much spam and bots and slop online that it’s going to be pretty unusable without technology to help you wade through the slop.”