The website looked like any small-town newspaper: a name that suggested a century of community coverage, a masthead, stories about city council meetings and high school sports. There was no byline on the articles, no address for the newsroom, and no evidence that a single human had written any of the words. The site was one of dozens, perhaps hundreds, in a network of AI-generated local news operations exposed in an investigative report this week, and it represents the newest chapter in the pollution of the information ecosystem.
The Verge’s investigation detailed how the network works. Automated systems scrape real news from legitimate outlets, rewrite the stories with AI, attach plausible-sounding headlines and publish them on sites designed to look like established local media. The sites carry advertising, which generates revenue, and they rank well in search results, which generates traffic. The result is a business model that monetizes trust it never earned, at the expense of the local journalists who actually cover the towns these sites pretend to serve.
The economics explain the epidemic. Local news has been collapsing for two decades: researchers say more than half of U.S. counties now have little or no regular local news coverage, and thousands of newspapers have closed or merged since 2005. The ad revenue that once supported reporters has migrated online, where it is captured by whoever ranks first in search. AI made it possible to fill the vacuum at near-zero cost — no reporters, no offices, no fact-checking, just a script and a domain name.
The fake sites do more than siphon advertising dollars. They spread misinformation under the cover of local credibility. Investigators found fabricated crime reports, invented community events and recycling of old stories presented as new — the kind of content that can shape public perception in small towns where there is no real newspaper left to correct the record. In emergencies, the stakes are higher: a fake weather warning or a fabricated public notice, dressed as local journalism, can do real harm.
The operators are hard to pin down. The sites hide behind privacy services, offshore registrars and shell companies, and the people behind them rarely leave a trail. Some appear to be run by content farms that operate hundreds of sites at once, in multiple languages. Others look like individual experiments in ad arbitrage. Law enforcement and regulators have begun to take notice, but the scale of the problem outruns the speed of enforcement.
Regulators have started to push back. The Federal Trade Commission has pursued actions against companies that use AI to generate fake content at scale, and state attorneys general have opened inquiries into fraudulent news sites. Google, whose search rankings the sites depend on, has tightened its policies against content that misrepresents itself, including rules aimed at so-called reputation abuse. The platforms say they are removing bad actors, but the operators adapt quickly, and each enforcement action seems to spawn a new generation of sites.
The pattern is not limited to the United States. Investigators have found similar networks in Europe, Latin America and Asia, targeting communities where local media is thin and verification is hard. The phenomenon has a name in the research literature — information laundering — and it is spreading along the same routes that once carried spam and disinformation: wherever there is money in attention and no one checking the source.
For the legitimate local press, the AI news farms are a double injury. They steal readers and revenue, and they dilute the credibility of the profession by imitation. Journalists’ groups have begun publishing guides to spotting AI-generated local sites — checking for fake bylines, boilerplate language, missing contact information — and some states have passed or proposed laws requiring disclosure when content is generated by AI. The disclosures, where they exist, have proved hard to enforce.
The deeper problem is structural. The conditions that made the fake sites profitable — the decline of local journalism, the dominance of algorithmic distribution, the low cost of content creation — are the same conditions that define the modern internet. AI did not create the vacuum in local news; it just found a way to fill it faster and more cheaply than anyone expected. Killing the sites would leave the vacuum intact, and something else would fill it.
The investigation ends with a question the industry has not answered: what is local news for, and who will pay for it? The AI networks have demonstrated that there is money in the appearance of journalism. The harder problem is whether there is money in the real thing — and whether communities, regulators and platforms can build defenses that prefer accuracy over imitation. The answer, when it comes, will determine whether the next decade of local news is written by reporters or by machines imitating them.
