Twitter has long been the platform that’s attracted the most controversy for “shadowbanning” — the practice of covertly suppressing a user’s content without their knowledge. But thanks to revelations in a recent Project Veritas leak, Twitter’s shadowbanning crown might have been stolen by Facebook.
The leak, which includes a company presentation on tackling “trolls” on Facebook, reveals a host of covert tactics that the social network proposes to use as means of censoring undesirable users without alerting them to the fact that they’ve been censored.
The presentation proposes the use of *forced glitches* like frequent “auto-logout[s]” and the failed upload of comments. These “special features” would be *triggered “leading up to important elections.”
It also describes a line of code called ActionDeboostLiveDistribution, which Facebook adds to unwelcome videos if it wants to “deboost” them, i.e, limit their visibility on the platform.
The emphasis on pre-election crackdowns confirms a pattern of behavior on the part of Facebook. Over the past two years, the social network has launched massive censorship waves ahead of elections around the world, including in France, Germany, and Brazil — in the case of Facebook-owned WhatsApp — Brazil.
More recently, Facebook launched a censorship wave against anti-establishment pages on its platform ahead of the 2018 U.S. midterm elections. The suppression included a decision not to allow a campaign ad about illegal immigration from the Trump campaign, and a pro-life advert from a campaign supporting Sen. Marsha Blackburn (R-TN)’s election bid.
The use of auto-logouts and other “forced glitches” as a tool of covertly interfering with a user’s experience also lines up with recent allegations made by Donald Trump Jr’s followers on Instagram, a Facebook-owned platform.
The Instagram users complained that their accounts had automatically unfollowed Don Jr, President Trump, and other accounts associated with the political right — all without their knowledge. The users also reported that they were being prevented from liking certain posts from right-wing and Republican figures.
One of the other remarkable details of the leak is that the tech giant’s covert suppression isn’t targeted at fringe individuals, but at mainstream conservatives like BlazeTV host Lauren Chen and Steven Crowder. Even independents like Mike Cernovich were caught in the mix.
According to the leak, Facebook has also developed an algorithmic classifier that automatically detects “trolling” behavior. There’s just one problem — the terms that Facebook has taught its algorithm to add to a user’s “troll” score include anodyne Internet-speak like “IRL” and “normie.”
One thing the social network users can take away from this leak — if you’re conservative and being automatically logged out or prevented from commenting ahead of a major election, it’s probably not an accident.