Dozens of News Sites are Using AI Chatbots to Churn Out Questionable Content

As far as synthetic, AI-generated content goes, we might just be at the tip of the iceberg.

In new report, NewsGuard, a firm that measures the credibility of online news sources, identified a whopping 49 websites producing content that appeared to be either mostly or entirely AI-generated in the month of April alone.

Designed to mimic the structure of real news sites, the scale of content produced by these sites — in addition to the quality and lack of transparency involved — is alarming, and might offer a glimpse into what an internet flooded with low-quality synthetic content might just look like.

Plastered with advertisements and riddled with bland, error-laden writing, these websites paint a fairly dismal picture of the world wide web's future, where misinformation-fueling, AI-generated content further muddies the lines between reality and fiction.

The scale of the operations identified by NewsGuard is immense.

"The websites, which often fail to disclose ownership or control, produce a high volume of content related to a variety of topics, including politics, health, entertainment, finance, and technology," reads NewsGuard's report. "Some publish hundreds of articles a day. Some of the content advances false narratives."

According to the report, the generic titles of these many sites — GetIntoKnowledge.com, HistoryFact.in, and BestBudgetUSA.com, to name a few — and the fact that they're plastered with advertisements suggests that they're designed to generate programmatic revenue through ads and affiliate links.

But the websites in question are chock full of glaring red flags. Author bylines, for example, are generally either vague or just nonexistent. Some of these sites, like HarmonyHustle.com, have gone as far as to list human-sounding authors like Alex and Tom. These bylines sometimes even feature their own profile pictures, which were likely AI-generated, according to NewsGuard.

Elsewhere, the writing on these content farms is generally both boring and repetitive, hallmarks of AI-generated content. It's also often filled with blatant falsehoods. For example, in one particularly egregious instance, a site called CelebritiesDeath.com claimed that president Joe Biden had died on April 1, 2023.

As it turns out, the site's listed source was an April Fool's joke tweet from a conservative Twitter personality.

Worse yet, CelebritiesDeath.com's dubious claim was even followed by OpenAI's misinformation disclaimer — a likely sign that no human had laid an eye on this particular post.

NewsGuard found similar in-text OpenAI disclaimers in at least one article on each of these 49 sites.

So who's behind these operations? Only 29 of these 49 sites listed contact information. Of those 29, only two confirmed that they have used AI in some capacity. As for the other 27, two declined to respond, 17 failed to respond at all, and eight had listed invalid email addresses.

Of those who responded was a person named Adesh Ingale, who identified himself as the founder of a site dubbed GetIntoKnowledge.com. But while the site features head-scratching and extremely incoherent headlines, Ingale assured to NewsGuard that his site only uses automation when it's "extremely needed."

"And yes they are 100 percent facts checked [sic] so that no false information is created," Ingale told NewsGuard. "As a world [sic] is growing towards digital and automation era we have introduced some automation softwares in our work but the results getting out of it are 100 percent original and regional facts based [sic]."

GetIntoKnowledge.com is "published manually under human supervision," Ingale added. "We are the new age of providing knowledge to each and every corner."

Of course, it might be easy to write off these sites as avoidable content garbage dumps. But while humans might be able to mentally flag sites like this, the real concern is that other AI systems, like Microsoft's Bing AI and Google's Bard, might not be able to — after all, they've already been known to volley misinformation back and forth.

In short, the report is concerning, to say the least. Some of the websites in question have already amassed massive followings on social media sites like Facebook, meaning that we should expect to see more of these sites crop up in the coming months.

More on AI: Microsoft Released an AI That Answers Medical Questions, but It’s Wildly Inaccurate