I Worked in Trust & Safety. Rose McGowan’s Twitter Suspension Makes No Sense
The Internet lit up with outrage after actress Rose McGowan, who has been leading the charge online against Harvey Weinstein and Hollywood harassment, had her account put on a 12-hour temporary suspension. Every way you look at it, this decision was a terrible one.
For more than a year, I worked on the Trust & Safety team at Tumblr. This is the department that deals with potential violations of Terms of Service, especially those that have a legal element. We were directly overseen by the General Counsel and his deputy attorney, and any decision that involved a gray area would be escalated up the chain to have a formal judgment made. When it came to celebrities, politicians or anyone who was designated a “VIP,” the issue was escalated as a matter of course. Which makes McGowan’s suspension on Twitter mind-boggling to me.
Does Twitter have none of these same practices in place—or are the decision-makers there just completely incompetent?
The Washington Post‘s article on McGowan’s suspension initially suggests that she could have fallen victim to the company’s abuse-fighting algorithms, either because of reports against her or by tripping something automatic:
Twitter introduced the 12-hour account limitation earlier this year. This particular punishment does not rely entirely on user reports of possible rule-breaking but instead can also be triggered automatically. If an account is tagging others in potentially abusive tweets or firing off a great deal of similar tweets flagged by the system as potentially abusive, those can be factors that prompt a suspension, the company said at the time.
However, as the Post further notes:
But like many of Twitter’s anti-abuse practices, there’s a lot of murkiness here to the enforcement. Even as the platform takes more aggressive steps to fight harassment and abuse, the actual enforcement of the rules designed to protect its users remains inconsistent.
Twitter remains a platform where everyday users still struggle with insults, death and rape threats, and are frequently told that the abusing accounts did not violate the network’s TOS and that Twitter’s hands are tied. Considering McGowan’s Tweets in recent days have attempted to hold the powerful to account for their crimes or the knowledge of those crimes, her account’s suspension could not have come at a worse time for Twitter—and for us.
Twitter is under fire daily for not doing enough to protect its users—particularly women—as well as continuing to enable Donald Trump’s account. They’ve explained that Trump’s threats of violence against North Korea don’t violate their TOS because of their “newsworthiness.” How are McGowan’s Tweets—and her ongoing ability to Tweet—anything less than newsworthy? That she was so unceremoniously silenced is probably one of the worst Trust & Safety decisions I’ve ever witnessed.
The Post does point out that one now-deleted Tweet on McGowan’s timeline contained an image of an email and an email signature that revealed personal information (a phone number) about the sender. This could be construed as a privacy violation, and the image most certainly could have been removed with a warning sent to McGowan’s account.
Generally, however, in my experience, information like this that was not distributed with the direct intent to dox the subject is already a privacy violation gray area. And even if the decision could rightfully be made to remove that single Tweet, the suspension of McGowan’s high-profile account at this moment in time continues to be an indefensible position.
Since I began writing this, Twitter’s team has publicly responded, claiming that McGowan’s account was locked because of the phone number Tweet and pledging to do better at communicating next time.
We have been in touch with Ms. McGowan’s team. We want to explain that her account was temporarily locked because one of her Tweets included a private phone number, which violates our Terms of Service. 1/3
— Twitter Safety (@TwitterSafety) October 12, 2017
Twitter is proud to empower and support the voices on our platform, especially those that speak truth to power. We stand with the brave women and men who use Twitter to share their stories, and will work hard every day to improve our processes to protect those voices. 3/3
— Twitter Safety (@TwitterSafety) October 12, 2017
This is, quite frankly, a load of horseshit. That it took the team an entire night and a considerable portion of this morning to reverse the decision and unlock the account, as well as issue a comment, is unconscionable from both the Trust & Safety and the PR perspective. So many balls were dropped here that it is raining balls.
Trump threatens war and the first amendment and tweets gifs of him assaulting people and targets private citizens but he’s still here.
— diane elyssa (@dianelyssa) October 12, 2017
Any Trust & Safety agent worth their salt would have immediately escalated any report concerning McGowan straight up to the highest ranks at Twitter. If it was decided therein that McGowan’s account should be temporarily suspended while entire troll armies of abusers continue to roam freely, the people responsible for this decision should no longer have jobs in decision-making.
McGowan was using the platform and her high profile to provide a voice for the voiceless at an extremely fraught time, and there’s nothing more “newsworthy” than that. And if the initial suspension was the result of an algorithmic trigger, it still should have been immediately reversed—especially so when the wall-to-wall media coverage began.
Twitter, you have a lot to answer for. Saying that you will be “be clearer about these policies and decisions in the future” is not enough. First, you need to take steps to protect your users from abuse, and you need to actually make decisions on those users’ behalf that make sense. This was a travesty of Trust & Safety.
(via The Washington Post, image: Shutterstock)
Want more stories like this? Become a subscriber and support the site!
—The Mary Sue has a strict comment policy that forbids, but is not limited to, personal insults toward anyone, hate speech, and trolling.—