Facebook has been promoting hundreds of AI-generated spam pages on its platform, which have recently gone viral, according to a report by 404 Media. The outlet found that most of these page creators were from India, Vietnam and the Philippines, and had learned how to make bizarre, AI-generated images from YouTube tutorials. 

Facebook’s recommendation algorithm rewarded these AI-generated spam pages which were posting dozens of times in a day. In turn, the creators were paid directly by the social media platform via its Creator Bonus Program, which pays people who post viral content. 

Renee DiResta of Stanford’s Internet Observatory and Josh A. Goldstein of Georgetown’s Center for Security and Emerging Technology analysed audience interactions with these viral AI content pages saying that they now accounted for hundreds of millions of interactions and were shown on the feeds of users who didn’t even follow these pages. 

Last week, a series of AI-generated images of Jesus melded with a crustacean had gone viral across the platform leaving people nonplussed. 

(For top technology news of the day, subscribe to our tech newsletter Today’s Cache)

From the comments made under these posts most users exposed to such images had no clue about the usage of AI image generation tools. 

The publisher spoke to multiple content creators of these viral pages, one of who shared that Facebook had been paying good performance bonuses on these types of content now. He added that it was easy to upload 50 or even 100 posts with these AI-generated images and make a decent amount of money. 

Another part of their business model was to sell courses on YouTube around how to generate revenue from these images on Facebook and earn ad revenue as well. 

A former Meta employee who spoke to the outlet also noted that Facebook is usually aware of this content but had laid off most of its content moderation employees leaving a leaner team to handle such large-scale issues. 

Earlier in April this year, The Washington Post reported that Meta was planning job cuts for its Oversight Board, an independent body that was overseeing the company’s content moderation decisions.