Facebook tackles the spread of deepfakes and troll farms

Facebook has taken down a troll farm, fake news spreaders and fake image creators as part of its latest moderation efforts.

The company’s latest Coordinated Inauthentic Behavior report, released this week, lists Facebook’s most recent efforts to reduce coordinated inauthentic behavior on the network.

According to the March report, Facebook investigated a “long-standing” troll farm and eliminated it.

A troll farm operated from Albania

Members of this troll factory primarily targeted an Iranian audience and are said to have ties to the Mojahedin-e Khalq (MEK) organization, an armed resistance group made up of several thousand members. The MEK was exiled to Albania in the 1980s and today appears to operate a network of genuine and fake accounts to disseminate information critical of the Iranian government and praising the activities of the MEK.

According to Facebook, sharing of content related to this organization peaked in 2017 and 2020 through three separate groups, but the majority of the group’s efforts to grow its audience have failed.

“Most of its accounts were managed by operators in Albania who regularly shared the technical infrastructure,” notes the company. “This means that the same operator could manage several accounts and, conversely, that several operators could manage the same account. These are some of the characteristics of what is known as a troll farm – a physical place where a collective of operators share computers and phones to jointly manage a pool of fake accounts as part of a trading operation. affecting. “

Facebook targets antagonistic generative networks

In addition, Facebook is tackling deepfakes, images generated by the application of artificial intelligence. While the company began suppressing fake images three years ago, antagonistic generative networks are now using deepfakes to masquerade as independent news outlets and investigative journalists.

After reviewing research provided by FireEye on a GAN network located in Spain and El Salvador, the company removed accounts and pages that posted municipal election information at a “spam-like” rate. Two other networks were also deleted, bringing the total number of deletions made by Facebook to seven operations using AI-generated images.

The social media giant has also documented its usual disruption of inauthentic networks. A total of 14 CIB operations were disrupted in March, originating from countries such as Argentina, Egypt, Israel, Mexico and Georgia, which led to the deletion of more than 1,100 accounts, 255 pages and 34 groups.

Last month, Facebook said it had successfully detected and dismantled a Chinese network of attackers using the platform to distribute malware. The operators, believed to be part of the Earth Empusa or Evil Eye groups, were using fake profiles to target journalists and activists.

Source: .com

Back to top button

Adblock Detected

Please consider supporting us by disabling your ad blocker