Twitter’s secret “Project Guardian” aims to protect controversial characters and users with large numbers of followers from an attack by trolls and haters, according to a report by Bloomberg. The platform is said to maintain a list of thousands of users who Twitter considers at high risk of harassment, which includes musicians, professional athletes, journalists and other particularly important users, even if only for now. .
Bloomberg Note that when Twitter receives a report of an abusive message linked to an account on the list, the Twitter content moderation team will respond to that report more quickly than any others it has listed. The idea behind the program is that Twitter can prevent the spread of harmful content, as well as keep content that is important to tweeters and less likely to report harassment on the platform.
Yoel Roth, Twitter Site Integrity Officer, said Bloomberg that a variety of users make up its Project Guardian list, and that they don’t have to be celebrities. Users who are caught up in viral Twitter drama can also be included, but usually only temporarily. After the 15 seconds of fame has passed, Twitter will remove that user from Project Guardian, while others will have a more permanent spot on the list. Like Bloomberg States, a user can be added to the program when a Twitter employee notices that they are seeing large volumes of hate messages, even if that user is not aware of it. On the other hand, a high profile user may even ask their manager to personally ask Twitter to offer more protection.
According to Bloomberg, some users registered with Project Guardian in the past include makeup artist James Charles, Egyptian activist Wael Ghonim, and former US Food and Drug Administration commissioner Scott Gottlieb. Twitter also used the program to protect reporters covering controversial topics, such as the January riots or 8chan.
In addition to Project Guardian, Bloomberg has found that Twitter uses several other factors to prioritize its responses to user reports. This includes the impression of a message, the number of subscribers of the user in question, as well as whether the reported tweet is actually dangerous or not. There is no indication that a particular incident triggered the creation of the program, but Bloomberg says it might have been around for two years or more.
Like Bloomberg points, Project Guardian doesn’t just protect users; it also protects Twitter from bad public relations. In March, Twitter was criticized for its inability to tame the bullies who bombarded former model Chrissy Teigen. The trolls targeted Teigen with a baseless claim that she was part of a conspiracy theory involving a celebrity pedophile ring. She quit Twitter following the bullying (although she has since returned). Critics argue that Twitter could have done more to protect Teigen, although Teigen herself has said she does not blame the platform for the barrage of abuse.
Right before this reveal, Twitter made some swift changes to the platform. Within a single day, Twitter acquired the Quill messaging platform, launched a test to change the reporting process, began piloting individual content warnings, and began experimenting with a TikTok style “For you” tab. The wave of announcements came in the first week or so after former Twitter CTO Parag Agrawal replaced Jack Dorsey as CEO. In an interview shortly after the Agrawal takeover, he said he hoped to speed up Twitter’s “execution” and streamline its operations.