The most high-profile users of Twitter Inc. – those with high followers or high profile – often receive heightened protection from the social network’s content moderators as part of a secret program that tries to limit their exposure to trolls and bullies.
The internal program, codenamed Project Guardian, has a list of thousands of accounts most likely to be attacked or harassed on the platform, including politicians, journalists, musicians and professional athletes. If someone flags offensive posts or messages related to these users, the reports will be prioritized by Twitter’s content moderation systems, which means the company will review them faster than other reports in the queue.
Twitter says its rules are the same for all users, but Project Guardian makes sure that potential issues related to celebrity accounts – those that could break out into viral nightmares for users and the company – are addressed before complaints from people who don’t do this. Part of the program.
This VIP group, which most members don’t even know is a part of, is designed to remove abusive content that could have the greatest reach and is most likely to be spread on the social media site. It also helps protect these celebrity users’ Twitter experience, making them more likely to keep tweeting – and potentially less likely to publicly complain about abuse or harassment issues.
“Project Guardian is just the internal name for one of many automated tools we use to identify potentially abusive content,” said Katrina Lane, vice president of the Twitter service organization, who operates the program. “The techniques used are the same that protect everyone on duty.”
The list of users protected by Project Guardian changes regularly and doesn’t just include famous users, according to Yoel Roth, Twitter’s head of website integrity. The program is also designed to increase protection for people who inadvertently come into the spotlight as a result of a controversial tweet or who were suddenly targeted by a Twitter mob.
That means some Twitter users will be temporarily added to the list while they have the world’s attention; others are almost always on the list. “The reason for the existence of this concept is the ‘person of the day’ phenomenon,” says Roth. “And based on that, there are some people who are the ‘person of the day’ most days and Project Guardian would be one way to protect them.”
The existence of the program raises an obvious question: if Twitter can protect some of its most visible users – or those who suddenly became famous – faster and more efficiently, why couldn’t it do the same for all the accounts on the receiving end of? ? Bullying or Abuse?
The short answer is yardstick. With over 200 million daily users, Twitter has too many abuse reports to deal with all of them at once. This means that reports are prioritized based on various data points, including how many followers a user has, how many impressions a tweet receives, or how likely it is that the tweet in question is abusive. Adding an account to Project Guardian is just one of those signals, although people familiar with the program believe it is a strong one.
Roth said the distinction can’t apply to everyone, or it would mean there is no point in having a list.
“If the list gets too big, it’s no longer a signal,” he added. “We really want to focus on the people who are getting exceptional or unprecedented notoriety at any given moment … this really focuses on a small subset of the accounts.”
Project Guardian was used to protect users from a number of different occupations. YouTube star and makeup artist James Charles was inducted into the program earlier this year after being harassed online. Egyptian internet activist Wael Ghonim was also part of the Project Guardian, as was former Commissioner for the U.S. Food and Drug Administration Scott Gottlieb, who often tweeted about COVID-19 vaccines. The program also included journalists – even news interns – writing on topics that could lead to nuisance, such as the January 6 riot in the US Capitol.
Twitter has used Project Guardian to protect its own employees, including Roth. After the company first reviewed then-President Donald Trump’s tweets in May 2020, Roth was selected by Trump and his supporters as the staff member behind the decision, which led to attacks and death threats. Roth, who wasn’t actually the person who made the call, said he was temporarily added to the Project Guardian’s list at the time.
“Suddenly I became much more famous than the day before,” explained Roth. He said he was removed from the program after the harassment subsided.
Accounts are added to the list in a number of ways, including following the recommendation of a Twitter agent who has witnessed an attack on a user and requests additional protection. In some cases, the manager or agent of a famous Twitter user will reach out to the company and ask for additional protection for their customer. Social media managers at news agencies have also called for additional protection for their colleagues who write high profile or controversial stories. Users who participate in the program may not know they are getting extra attention.
“We see it as, who are the people who we know have been victims of abuse or who are predicted to be likely victims of abuse?” Said Roth.
Twitter said it’s getting better and better to automatically detect abuse and harassment, which means it doesn’t have to wait for a user to report an issue before sending it to a human moderator. The company says its technology now has 65% of the abusive content it removes, marks or prompts people to delete before a user ever reports it.
It’s not clear if there was an event or incident that sparked Project Guardian, although it has been around for at least a couple of years, people familiar with the program said.
In recent years, Twitter’s image has suffered when high-profile users publicly criticize the service – or abandon it altogether – for failing to combat abuse and harassment. It is especially common among famous women. Model Chrissy Teigen, singer Lizzo, actor Leslie Jones and New York Times journalist Maggie Haberman have all publicly resigned from the service after being inundated with negative tweets and news. (They have all returned now.)
More recently, however, it seems less common for celebrities to call Twitter out for constant harassment, and some people familiar with the company believe Project Guardian is a reason for this.
Twitter’s program is another example of the differential treatment social media apps offer certain outstanding users and accounts. A September research report by the Wall Street Journal found that Meta Platforms, which includes Facebook and Instagram, granted some celebrity users special exceptions to some of their rules, leaving content from those individuals that would have been flagged or removed by others.
Twitter officials firmly believe that Project Guardian is different and that all users on its platform are bound by the same rules. Reports related to users who are part of Project Guardian are scored the same as any other content report – the process is usually faster.
While Twitter’s rules can apply to everyone, the penalties for violating these rules are not always the same. For example, world market leaders have more leeway when breaking the rules of Twitter than most of its users. Twitter and Meta have also spent years cultivating relationships with high profile users and creating teams to help these people use their products and provide hands-on assistance when needed. In 2016, Twitter stopped hiring a small group of celebrity users with ads to improve their experience.