Left-wing and mainstream technology news outlets ignored Project Veritas’ Twitter censorship exposé, which revealed Twitter employees admitting to censorship against conservatives this week.
TechCrunch, the Verge, Mashable, TechRadar, CNET, and Ars Technica all ignored the news, which was only picked up by conservative outlets.
The Project Veritas investigation, which secretly recorded several former and current Twitter employees, including Twitter Trust and Safety Policy Manager Olinda Hassan, revealed the employees discussing and boasting about the ways they censor conservatives, pro-Trump accounts, “shitty people,” and users who express interest in God, guns, and America.
As reported by Breitbart News’ Allum Bokhari on Thursday, “The latest video from James O’Keefe’s Project Veritas shows Twitter employees appearing to admit that the platform has ‘shadow banned’ users in the past, and applies its rules and censorship algorithms unevenly in an effort to rid the platform of Trump supporters and conservatives.”
Shadow banning is the act of banning a user from a platform without them realizing. Shadow banned users will typically be removed from the public eye, so other users cannot see or interact with them, however since the user is not alerted to the shadow ban, they will not understand why their posts suddenly receive no engagement.https://archive.fo/dTUBN
Twitter direct messaging engineer Pranay Singh admitted to mass-banning accounts that express interest in God, guns, and America, during a Project Veritas investigation.
“Just go to a random [Trump] tweet and just look at the followers. They’ll all be like, guns, God, ‘Merica, and with the American flag and the cross,” declared Singh, who was secretly recorded by Project Veritas reporters. “Like, who says that? Who talks like that? It’s for sure a bot.”
After being asked whether he could get rid of the accounts, he replied, “Yeah. You just delete them, but, like, the problem is there are hundreds of thousands of them, so you’ve got to, like, write algorithms that do it for you.”
“So if there’s like ‘American, guns,’ [in the account bio] can you write an algorithm to just take all those people out?” asked one undercover reporter.
“Umm, yeah, it’s actually how we do it,” Singh replied. “You look for ‘Trump,’ or ‘America,’ or any of, like, five thousand, like, keywords to describe a redneck, and then you look, and you parse all the messages, all the pictures, and then look for stuff that matches that stuff… You assign a value to each thing, so like Trump would be .5, a picture of a gun would be like 1.5, and if the total comes up above a certain value, then it’s a bot.”
An undercover reporter then asked Singh whether the “majority of the algorithms” are “against conservatives or liberals,” to which he responded, “I would say a majority of it are for Republicans, because they’re all from Russia, and they wanted Trump to win.”
“So you would mostly just get rid of conservatives?” asked the reporter.
“Yeah,” he replied.https://archive.fo/7iKUt