It was found that there are greater than 50 ‘content farms’ sites that abuse generative artificial intelligence (AI) comparable to ‘ChatGPT’ to mass-produce fake news.

Citing a report by NewsGuard, a fake news monitoring group, Bloomberg on the first (local time) identified dozens of reports sites using AI chatbots, and 49 of them were found to be “News Live 79” or “Every day Business Post.” It was reported that they were spreading fake breaking news under their names, lifestyle suggestions and star news.

Last month, an article titled ‘President Biden’s death’ appeared on a site called ‘Celeb Death’, and a site called ‘T News Network’ posted false facts that 1000’s of individuals were killed within the war in Ukraine.

These content farms have the aim of attracting clicks with sensational articles and winning advertisements. It is working in several languages, including English in addition to Portuguese and Thai, and has been confirmed to be spread world wide. As well as, 24 sites, half of which were surveyed, used Google ads to generate revenue.

“To do something like this using an AI model known to control facts is fraud under the guise of journalism,” said NewsGuard co-CEO and former Wall Street Journal publisher Gordon Prowitz. With the intention to prevent this, he identified that more attention needs to be paid to constructing guardrails for AI models.

In an announcement from a spokesperson, Google said it “focuses on the standard of content reasonably than the way it was created, and blocks or removes ads when it detects violations.” It also announced that it had actually removed some advertisements in response to inquiries.

Noah Gian, an associate professor at Bentley University in Syracuse, said, “Prior to now, labor costs were involved in creating fake news, but now it’s free and easier and faster. Evidently more content factories shall be automated.”

Reporter Lim Dae-jun ydj@aitimes.com