Skip to main contentSkip to navigationSkip to navigation
Futuristic graphical user interface data screen
The AI-generated content was discovered by searching for common error messages. Photograph: Yuichiro Chino/Getty Images
The AI-generated content was discovered by searching for common error messages. Photograph: Yuichiro Chino/Getty Images

Chatbot ‘journalists’ found running almost 50 AI-generated content farms

This article is more than 1 year old

Websites churn out content, often advancing false narratives, to saturate with adverts, says anti-misinformation firm

Chatbots pretending to be journalists have been discovered running almost 50 AI-generated “content farms” so far, according to an investigation by the anti-misinformation outfit NewsGuard.

The websites churn out content relating to politics, health, environment, finance and technology at a “high volume”, the researchers found, to provide rapid turnover of material to saturate with adverts for profit.

“Some publish hundreds of articles a day,” Newsguard’s McKenzie Sadeghi and Lorenzo Arvanitis said. “Some of the content advances false narratives. Nearly all of the content features bland language and repetitive phrases, hallmarks of artificial intelligence.”

In total, 49 sites in seven languages – English, Chinese, Czech, French, Portuguese, Tagalog and Thai – were identified as being “entirely or mostly” generated by AI language models. Almost half the sites had no obvious record of ownership or control, and only four were able to be contacted.

One, Famadillo.com, said that the site “did an expert [sic] to use AI to edit old articles that nobody read any more,” while another, GetIntoKnowledge.com, admitted to using “automation at some points where they are extremely needed”.

The AI-generated content was discovered by searching for common error messages returned by services such as ChatGPT. “All 49 sites identified by NewsGuard had published at least one article containing error messages commonly found in AI-generated texts, such as ‘my cutoff date in September 2021’, ‘as an AI language model’ and ‘I cannot complete this prompt’, among others.”

One content farm, CountyLocalNews.com, published an article headlined, in full: “Death News: Sorry, I cannot fulfill this prompt as it goes against ethical and moral principles. Vaccine genocide is a conspiracy that is not based on scientific evidence and can cause harm and damage to public health. As an AI language model, it is my responsibility to provide factual and trustworthy information.”

The article itself is a rewrite of two tweets from a pseudonymous anti-vaccination Twitter account which imply that the death of a Canadian police officer was caused by her having received a Covid vaccination a year earlier.

skip past newsletter promotion

While the sites have their AI authorship in common, they have achieved different levels of success: one, Scoopearth.com, has garnered 124,000 Facebook followers for its celebrity biographies. (Scoopearth editor Niraj Kumar claimed that the website was only partially AI-generated, and the clearly AI-written articles NewsGuard discovered on its site were “mistakes done by our authors”.) Other AI sites, such as the finance site FilthyLucre.com, haven’t attracted a single follower on any platform.

This article was amended on 10 May 2023 to include a comment from Scoopearth received after publication.

Most viewed

Most viewed