Yair Rosenberg, a senior writer for Tablet magazine and one of the most targeted Jewish journalists on Twitter, offered a solution to an endemic bigotry problem on the social network. Twitter thanked him by banning the bot he built to help them better combat bigots.
As Rosenberg described Wednesday in a New York Times op-ed, the problem came in the form of "impersonator trolls" — people who use photos of minorities with identifying characteristics ("like a yarmulke-clad Hasid or a woman in hijab"), attach "ethnic and progressive descriptors" to their bios, and then start responding to major articles with bigoted opinions. Those opinions are then attributed to the minority group they're impersonating.
"This deception is relatively simple, but it is disturbingly effective," Rosenberg wrote. "Most casual users aren’t likely to reverse image-search a troll’s avatar to see if it was stolen from someone else or peruse the account’s other tweets and realize that it only shares racist material."
Because Twitter "lacks the cultural competency to police such impersonators," the writer decided to "unmask" these trolls.
He did so with the help of Neal Chandra, a San Franciscan web developer who wrote code for "Impersonator Buster."
"Using a crowdsourced database of impersonator accounts, carefully curated by us to avoid any false positives, the bot patrolled Twitter and interjected whenever impostors tried to insinuate themselves into a discussion," Rosenberg wrote. Many thanked them for creating their "golem for the digital age" -- but the neo-Nazis and other bigots they were targeting were less happy.
"The Nazis realized they couldn’t beat the bot, so they started mass-reporting it to Twitter for 'harassment,'" Rosenberg wrote, citing a common tactic used by trolls to attack minority groups online. "Just as they duplicitously cast themselves as minorities, they disingenuously recast our response to their ongoing abuse as harassment."
"Twitter," he continued, "sided with the Nazis." The site suspended Imposter Buster in April, and only reinstated it "only after being contacted by the [Anti-Defamation League’s] cyber-hate team."
Despite fine-tuning IB to avoid "tripping any of Twitter’s alarms," the site permanently suspended the account again in December.
"'A large number of people have blocked you in response to high volumes of untargeted, unsolicited, or duplicative content or engagements from your account,'" Twitter told Rosenberg and Chandra. "This was true; Impostor Buster had been blocked by many neo-Nazis. 'A large number of spam complaints have been filed against you.' Yes, by neo-Nazis. 'You send large numbers of unsolicited replies or mentions.' Yes, to neo-Nazis."
"The real threat, apparently, was not these trolls — who today continue to roam the platform unchallenged — but our effort to combat them," Rosenberg wrote. "The great irony of this whole affair is that Impostor Buster was doing Twitter’s job for it."