Twitter made attempts to tackle the bots and trolls on its platform after some were identified as helping sway elections and fueling political feuds. But, a new study found that Twitter still struggles to keep up with its abusive accounts.
Zubair Shafiq, a computer science professor at the University of Iowa and his graduate student Shehroze Farooqi conducted a 16-month study of 1.5 billion tweets. They found more than 167,000 apps are using Twitter’s API to automate bot accounts that spread millions of tweets tricking users into compromising their accounts, astroturfing, and other forms of spam.
A machine-learning approach developed by the researchers identified abusive accounts in far greater volumes and faster than Twitter itself does. For about 40 percent of the abusive applications checked, the researcher’s model was able to detect them at least a month before Twitter.
“We show that many of these abusive apps used for all sorts of nefarious activity remain undetected by Twitter’s fraud-detection algorithms, sometimes for months, and they do a lot of damage before Twitter eventually figures them out and removes them,” Farooqi told Wired.
The researchers had been sharing their results with Twitter for months, but the company hadn’t asked for further details.
Within the study’s abstract, the researcher’s noted they referred to a third-party Twitter application as “abusive” if it was in violation of Twitter’s rules. However, the results of the study aren’t too surprising, because Twitter has a long history of not responding to other types of abuse.
In 2018, an Amnesty report detailed online abuse against women on Twitter. The findings revealed Black women were disproportionately targeted by online abuse and were 84 percent more likely than white women to be mentioned in abusive or problematic tweets.
“Twitter’s failure to crack down on this problem means it is contributing to the silencing of already marginalized voices,” Milena Marin, Senior Advisor for Tactical Research at Amnesty International said.
Twitter has started banning accounts trying to evade suspensions for their abusive behavior and, in July 2018, announced plans to crackdown on spam accounts. The researchers told Wired that Twitter is moving in the right direction, but it’s clearly not enough.
