Artificial intelligence: Facebook pays for synthetic image spam pages

Facebook is promoting hundreds of spam pages containing images generated by artificial intelligence (AI), allowing fake content to be shared online in “huge volumes”, experts say.

Facebook's recommendation algorithm promotes these pages, which, thanks to AI, can publish several dozen synthetic images per day thanks to a bonus program for their creator, according to a report from 404 Media on Tuesday.

The more reactions and comments a post receives, the more rewards it gets, the tech-focused online media outlet continues.

” [Le programme de bonus] It offers creators the opportunity to earn money based on interactions with their publications,” we can also read on the website of Meta, the parent company of Facebook and Instagram.

According to researchers at the Internet Observatory at Stanford and the Center for Security and Emerging Technology at Georgetown, these viral image factories now have hundreds of millions of interactions and appear in the news feeds of users who don't follow them.

Most of the creators of these pages are said to be from India, Vietnam and the Philippines, and according to 404 learned to create the strange AI-generated images through tutorials on YouTube.

Independent media interviewed several content creators of these viral pages, who said Facebook pays good performance bonuses for this type of content.

But the danger of the rise of synthetic media shared and disseminated online is that “freely available and largely unregulated tools allow anyone to produce disinformation and false content on a large scale,” said a study by the research arm from German Media Deutsche. well

And based on the comments below the posts, many exposed users can't tell the difference between real images and those generated by AI, according to the report.

A former Meta employee who spoke to 404 Media said Facebook was aware of the content circulating on its platform, but reported a shortage of moderators from the latest wave of layoffs.

Last April, the Washington Post reported that Meta was planning job cuts at its Oversight Board, an independent body responsible for overseeing the company's content regulation decisions.

If the trend continues, experts predict that “up to 90% of online content will be artificially generated. [par l’IA] by 2026,” according to a report from Europol, the European police agency.