Trolls Are Paying Twitter to Promote Hate Speech — And There's Nothing Stopping Them
Wednesday morning, a sponsored Twitter ad imploring transgender people to commit suicide slipped through the cracks of Twitter's anti-hate content ad policy. It went live from a parody account mocking Australian feminist campaigner Caitlin Roper.
The parody account began spamming Roper's followers with sexually explicit content and solicitations.
It's surprisingly easy to promote hate speech on Twitter. Roper's troll said he put in just $25. Twitter ads cost as little as 50 cents each, which means the troll may have paid for about 50 "engagements" — retweets, replies and favorites.
Twitter bans ads relating to hate speech, sexual content and violence. Nonetheless, the Roper incident was the third time this month that trolls have succeeded in paying Twitter to promote hate.
Because of Roper's history of feminist activism, some suspected that Roper's attackers were followers of men's rights news site A Voice for Men. The blog, whose writers are outspoken critics of feminism, has targeted Roper for her work in the past.
However, the person behind the attack on Roper came from 4chan, the armpit of the anonymous Internet and the older, more chaotic brother of Reddit. (Warning: What you're about to read may be disturbing.)
In a post on 4chan's /pol/ board from May 13, a user detailed how he planned to buy the ad from Twitter and post it using Roper's name and profile photo. "I'm going to use Twitter Ads to send out 'kill yourself' messages to trannies," the user wrote. "Using Twitter Ads, you can send out promoted Tweets to millions of people, and you can target specific people to receive your Tweets. Twitter doesn't check Tweets beforehand, so I made a bunch of Tweets encouraging trannies to kill themselves, and I customized it so that trannies will be the ones receiving those Tweets."
He added, "Hopefully, I'll actually be able to get a few trannies to really kill themselves. The only good tranny is a dead one."
Then, on Wednesday, May 20, what looks to be the same user — he's the only one in both threads whose flair is the Jamaican flag — claimed credit for the prank and announced that the promoted tweet was a success. He urged fellow users, "Make sure to harass this feminazi whore for 'promoting transphobia.'" He explicitly stated that he hoped this caused people to kill themselves.
Other submissions on /pol/ by a user with the same Jamaica flair show support for hate groups on Reddit and anti-Semitic speech on Twitter, as well as sexist rants against "slutty" women who aren't "wife material." On Wednesday, he also claimed credit for promoting a tweet by a parody account of Guardian columnist Josh Bornstein, saying "My fellow Jews, we must EXTERMINATE Arabs."
(4chan is anonymous, so it's hard to say with certainty that these posts belong to the same person. And on 4chan, this type of sexist and racist content is not uncommon.)
Twitter has a trolling problem: In a response to Gizmodo's Kate Knibbs, a Twitter spokesperson wrote, "Twitter does not allow the promotion of hate content, including hate speech against a group based on sexual orientation or gender identity. ... Once this instance was flagged, we immediately suspended the account and stopped the campaign."
But the message here is troubling: Twitter is not screening potential ads for hate speech, instead relying on users to report offensive content.
Twitter is massive. Millions of people use it. To protect its users from harassment, failsafes and stronger monitoring tools need to be put in place for the social media giant to stop allowing things like this from happening again. Earlier this month, there was nothing preventing white supremacist Andrew 'weev' Auernheimer from using Twitter to promote his hate messages. Even more damning, Twitter got paid for it.
Twitter is reportedly gearing up to wipe porn off the network entirely, which would require kicking off about 10 million users, including spam accounts and sex workers alike. But it seems the site has a bigger problem on its hands if it's trying to clean up its act.