A French journalist (translate to english) traced more than 4k news websites built almost entirely using AI tools. Not spun content. Not repurposed PLR. Fully synthetic sites: headlines, stories, images, entire layouts. Allegedly the goal wasn't long-term credibility - it was short-term manipulation of Google Discover and content farm monetization.
Each site in the network followed a simple formula: Use AI to generate topical news content, run it through a style filter to mimic journalistic tone, and pair it with stock images or generated visuals. Then publish across a sprawling network of cloned WordPress instances, often with automated on-page SEO already baked in - H1s, metadata, alt tags, internal links, and Discover-friendly formatting. Many even included author bylines using fake names, complete with bios and profile pictures.
I don't condone such behavior, but having seen a couple dozen of these sites (for research purposes - hold my beer) I couldn't help but marvel at the beauty of the entire system. They simply poured in keyword buckets they wanted to target, keywords to stories listed from Google searches, and in less than a couple minutes a new legit looking website comes out the other end. When I say "legit looking" - here's the rub: I could not tell they were not real sites. The content was of sufficient quality to pass many blogs in the same spaces. Running the content through AI detectors threw up flags everywhere. So, the reality here, is this not a website problem, this is a Google problem. It is also an excuse for another salty journalist to carp about Google and AI.
There’s nothing particularly groundbreaking in the components — what’s notable is the level of integration. It’s the kind of system any technically savvy marketer could build in a weekend.
While the 4k may sound like a huge scale, it is the tip-of-the-iceberg for some of these systems. Several of these content farms own domain name registrars and are admins of TLDs. This gives the ability to 'wildcard' create domains at staggering scale. 4k, think more in terms of 400k in coming. This is all very deja-vu of when the Chinese groups started caching websites, reserving them to Google and getting higher rankings than the original website.
The underlying stack likely included ChatGPT or Claude for the writing, Midjourney or DALL·E for the images, a headless CMS or mass WordPress deployment script, and tools like Zapier, PhantomBuster, or custom scripts to automate publishing and indexing. Some networks added AdSense scripts and outbound affiliate links. Others were optimized purely for link farming — generating content around specific anchor text and auto-inserting backlinks across their own domains.
The takeaway isn’t panic about misinformation. It’s that content generation, SEO optimization, image creation, and publishing are now fully automatable at scale. If low-tier operators are chaining these tools together to build fake news empires, what are the possibilities for legit brands, agencies, and publishers who use the same methods with real oversight?
This isn't a wake-up call about content farms. It’s a signal about what’s now possible - the caliber of content - and what’s coming next in content production workflows.
As the CEO and founder of Pubcon and WebmasterWorld., Brett Tabke has been instrumental in shaping the landscape of online marketing and search engine optimization. His journey in the computer industry has spanned over three decades and has made him a pioneering force behind digital evolution. Full Bio
Visit Pubcon.com