A crowd-sourced study by a leading humanitarian group and global artificial intelligence company confirms that women around the world face abuse when using Twitter—and that without the social media platform's commitment to combating such treatment, pervasive online abuse will continue to have the effect of harming and silencing women, particularly women of color.
With the help of the artificial intelligence software company Element AI and 6,500 volunteers who sifted through more than 200,000 tweets sent to women in 2017, Amnesty International reported in their study, entitled "Troll Patrol," that women were mentioned in 1.1 million abusive or problematic tweets last year.
— Dina Bass (@dinabass) December 18, 2018
The tweets that the volunteers studied were sent to 778 female politicians and journalists, but that data was used to estimate that an abusive tweet is sent to a woman once every 30 seconds.
Women of color were about 34 percent more likely to face abuse on Twitter than white women, with black women particularly at risk—receiving 84 percent more abusive tweets than their white counterparts.
"With the help of technical experts and thousands of volunteers, we have built the world's largest crowdsourced dataset about online abuse against women," said Milena Marin, senior advisor for tactical research at Amnesty International. "Troll Patrol means we have the data to back up what women have long been telling us—that Twitter is a place where racism, misogyny and homophobia are allowed to flourish basically unchecked."
On Twitter, many women said they were not surprised by the findings, but expressed appreciation for a wide-ranging study that confirmed their own experiences online.
New research by @amnesty into abuse against women on Twitter found that 1.1 million "abusive or problematic" tweets were sent to women last year. That equates to one every 30 seconds on average. It's horrifying, but for most women, it's not surprising. https://t.co/ia8XlCeot8
— Rachel Thompson (@RVT9) December 18, 2018
So our suspicions are confirmed. According to Amnesty and @element_ai's crowdsourced data out today, women of colour are 34% more likely than white women to be abused on #ToxicTwitter. https://t.co/JdL8hxnWAw
— Shiromi (@blimundaseyes) December 18, 2018
"Amnesty's crowdsourcing is the most revealing data available for a problem that so many people know about but haven't been able to quantify," wrote Emily Dreyfuss at Wired.
In the tweets the volunteers studied, women across the political spectrum faced abuse, but left-leaning politicians were 23 percent more likely to receive threatening or abusive tweets. Meanwhile, right-wing female journalists were 64 percent more likely to be targeted on Twitter.
"Twitter's failure to crack down on this problem means it is contributing to the silencing of already marginalized voices." —Milena Marin, Amnesty InternationalThe study was borne out of Amnesty International's frustration with Twitter's dismissal of the issue of online abuse. CEO Jack Dorsey has repeatedly expressed his understandable view that the company aims to protect free speech rights and avoids banning users—but Amnesty took issue with the company's refusal to release "meaningful information about reports of violence and abuse against women, as well as other groups, on the platform, and how they respond to it," and conducted the study to demonstrate how many women are facing abuse on Dorsey's platform every day.
When the group shared the data with Twitter, the company only asked for clarification on how Amnesty defines a "problematic" tweet, "in accordance with the need to protect free expression."
Amnesty defines problematic tweets as containing "hurtful or hostile content, especially if repeated to an individual on multiple or occasions."
Such tweets "can reinforce negative or harmful stereotypes against a group of individuals (e.g. negative stereotypes about a race or people who follow a certain religion)" and "may still have the effect of silencing an individual or groups of individuals" even though they don't fit Twitter's definition of abuse.
"Troll Patrol isn't about policing Twitter or forcing it to remove content," Marin said. "We are asking it to be more transparent, and we hope that the findings from Troll Patrol will compel it to make that change. Crucially, Twitter must start being transparent about how exactly they are using machine learning to detect abuse, and publish technical information about the algorithms they rely on."
Given that women of color face disproportionate abuse on the platform, she added, "Twitter's failure to crack down on this problem means it is contributing to the silencing of already marginalized voices."