Cybersecurity researchers have detected a coordinated network of more than 200 AI-generated websites designed to exploit advertising networks and mislead users. The operation emphasize how generative AI is increasing digital fraud by making it less expensive and rapid to produce massive volumes of deceptive content, via AXIOS.
The investigation, performed by cybersecurity organization DoubleVerify, located that the websites were formed using of templated prompts fed into a large language model. Researchers named the operation “AutoBait.”
As per the report, the group deployed hundreds of “made-for-advertising” websites including low-quality, AI-generated articles paired with sensational headlines and slideshow-style content. These sites goal to seize advertising impressions instead of building legitimate audiences.
Exposed Prompts Reveal How the Operation Works
A main discovery in the investigation came from the operators themselves. The fraudsters by chance left the content-generation prompts fixed within the sites’ JavaScript code, offering a unprecedented glimpse into how AI tools power such schemes.
The prompt instructed the model to layout articles that might maximize engagement by emotionally charged content. The instructions highlighted dramatic storytelling and attention-grabbing headlines.
For example, the model was told to place “the most sensational or surprising factors — some thing that prevents someone mid-scroll” at the starting of slideshow-style articles. Headlines were needed to be “ultra-literal,” whilst the body text required to “inject real emotion (fear, anger, shock, remedy) into every paragraph.”
The prompt also directed the AI system to generate images that regarded authentic. Photos were needed to look “ultra-realistic,” as if they were “casually taken on a smartphone by real person.”
Advertising Fraud at Scale
The primary intention of these sites was not readership however monetization. The pages have been fulfil of advertising banners designed to generate sales from programmatic advertising and marketing networks. DoubleVerify evaluate that each article page price the operators less than $2.25 to generate using AI tools. With a couple of advert placements on every page, the operators may want to rapidly recover those costs through fraudulent advertising and marketing impressions.
This approach lets in scammers to exploit legitimate brands without their knowledge. Corporations that purchase digital advert placements may unknowingly fund fraudulent websites designed to manipulate users and inflate advert metrics.
Expired Domains Fuel the Network
Researchers also discovered that many of the websites had been built on earlier legitimate domain whose registrations had expired. By purchasing these domains, scammers inherited current traffic and domain authority.
Gilit Saporta, senior director of fraud and quality analysis at DoubleVerify, explained the strategy in comments shared with Axios. Some domains formerly hosted real blogs or informational sites, such as one related to a financial blogger who had passed away. The reuse of expired domains allows fraudsters to quick build networks that seem credible to both customers and marketing systems.
AI Is Lowering the Barrier for Online Scams
Online scams and advertising fraud have existed for years. Moreover, generative AI considerably reduces the cost and complexity needed to release large-scale operations. “AI is supercharging so many types of attacks, and we know that fraudulent actors are constantly going to be early adopters of any new, cool trick that they can use,” Saporta stated.
As AI tools preserve to enhance, cybersecurity experts warn that automated content generation may permit even more sophisticated deception campaigns.
