Instagram Fails to Address Hate Speech Against Women Politicians, Study Finds

Instagram is failing to crack down on hate speech against women politicians as elections near, advocacy group says

U.S. Vice President and Democratic presidential candidate Kamala Harris speaks during a campaign event in Eau Claire, Wisconsin, U.S., August 7, 2024. Instagram has failed to remove toxic comments directed at Vice President Kamala Harris and other leading female politicians from its app as the 2024 election nears, according to research from the Center for Countering Digital Hate. The non-profit advocacy is analyzing large internet platforms to see if they're properly monitoring their sites for hate speech. The report on Wednesday was based on an analysis of 560,000 comments on Instagram posts from five Republican and five Democratic women politicians with high levels of engagement. The politicians the group tracked included Harris, who's now the Democratic nominee for president, Rep. Alexandria Ocasio-Cortez (D-NY) and Rep. Jasmine Crockett (D-Texas) as well as Republican House members Marjorie Taylor Greene from Georgia and Lauren Boebert from Colorado. Of the comments posted between Jan. 1 and June 7, the researchers identified over 20,000 that were deemed to be "toxic" by Google's Perspective AI content moderation tool. The researchers then conducted a manual analysis and discovered 1,000 comments that "clearly breached the terms of Instagram," said CCDH CEO Imran Ahmed during a media briefing on Tuesday. "Our recommendations can be summarized quite simply as Instagram must enforce its policies designed to protect women in public life," Ahmed said during the briefing. "Organizations need to be better equipped to support women candidates experiencing abuse and to give them best practice on dealing with it quite often." Meta, Instagram's parent company, has repeatedly come under fire from lawmakers for failing to address the spread of hateful content across its family of apps and for its inability or unwillingness to crack down on harmful behavior. The attorney general of New Mexico has alleged in an ongoing lawsuit against Meta that the company is failing to protect underage users from predators and sexual exploitation. In prior election cycles, Facebook has also been a hub for the spread of misinformation and toxic content directed at political candidates. Some of the problematic comments captured by CCDH included statements like "make rape legal" and "we don't want blacks around us no matter who they are," the report said. One comment directed at Harris mocked her racial background, while another comment called for her sexual assault by President Joe Biden. The CCDH researchers then used Instagram's own content reporting tools to flag the 1,000 offensive comments it manually discovered. A week later, "Instagram had taken no action against 926 of them, equivalent to a failure to act on 93% of them," the report said. Meta said in a statement that it would review the examples that CCDH highlighted and will remove comments that violate company policies, but added that some content may be offensive but not in violation of its rules. The company also said that the Google AI tool that CCDH relied on for part of its research isn't always accurate. "We provide tools so that anyone can control who can comment on their posts, automatically filter out offensive comments, phrases or emojis, and automatically hide comments from people who don't follow them," Cindy Southworth, Meta's head of women's safety, said in a statement. "We work with hundreds of safety partners around the world to continually improve our policies, tools, detection and enforcement, and we will review the CCDH report and take action on any content that violates our policies." Regarding the racist comment directed at Harris, one of the CCDH researchers eventually received an Instagram notification that said the post "doesn't go against our Community Guidelines," the report said. The report also said that over one-fifth of the 1,000 offensive comments that the researchers flagged came from "'repeat offenders' who had posted abuse at least twice." The report on Instagram comes a few months after a California federal judge dismissed a lawsuit against CCDH by Elon Musk's X. The suit was filed shortly after the group published research showing an increase of hate speech in the aftermath of Musk's acquisition of the site formerly known as Twitter. Because of all the negative attention that's flowed in Musk's direction, Meta and CEO Mark Zuckerberg have escaped scrutiny of late and there's a perception that Instagram "has become a platform that people feel safe" using, Ahmed said. "Mark Zuckerberg has played a strategy of keeping his head down while X is acting as a lightning rod for a lot of the anger at the toxicity in public life and political discourse," Ahmed said. "We wanted to look specifically at that platform to see whether or not they're actually backing up some of their gloating of X's misfortunes with action of their own."

Instagram is failing to protect female politicians from abuse and anti-hate speech

Vice President Kamala Harris was one of 10 female politicians looked at in the Center for Countering Digital Hate study. Credit: Bloomberg / Contributor / Bloomberg via Getty Images In our first election cycle since Twitter's demise, Instagram is more important than ever. Despite being poised as the more respectable social media platform for candidates in comparison to Elon Musk's X and political target TikTok, Instagram isn't without its problems. According to the Center for Countering Digital Hate, Instagram is failing to protect female politicians from abuse on its platform. A new study found that one in 25 comments on female politicians' Instagram posts are "highly likely" to be toxic, and Instagram fails to act on 93 percent of reported abusive comments targeting female politicians. The CCDH is urging Instagram to transparently enforce its community guidelines and provide support for female politicians dealing with online abuse, as well as for lawmakers to hold social media companies accountable for the abuse on their platforms. "This sort of abuse can dissuade women from choosing politics and running again. While it doesn't necessarily move elections, it does signal women's place in political life," Imran Ahmed, the CEO and founder of the CCDH, said during a press briefing on Tuesday. Researchers at the CCDH selected 10 female incumbents from across the political spectrum running for office in 2024 and collected 560,000 comments on their Instagram posts dating from January to June 2024. Politicians were selected based on high digital engagement and comments studied include those on posts by Vice President Kamala Harris, Congresswoman Alexandria Ocasio-Cortez, Congresswoman Marjorie Taylor Greene, and Congresswoman Marsha Blackburn. Utilizing Google's Perspective AI tool, which is trained to identify toxic text, it found over 20,000 "rude, disrespectful, or unreasonable comments." While it's unclear whether or not all 20,000 comments violated Instagram's community agreement, the study went a step further by having two researchers sift through the comments to identify the 1,000 most vile comments and report them on Instagram. A week later, 926 of the reported comments remained on the platform. These comments included sexist and racist remarks, and death and rape threats, all of which clearly violate Instagram's policies. Seventy-seven percent of the comments included gendered abuse such as "bitch," "rape," and "whore." Instagram boasts that it "removes content that contains credible threats or hate speech" and "content that targets private individuals to degrade or shame them," but its lack of response to the majority of the comments in this study suggests otherwise. Additionally, those who report comments on Instagram are given very few details on why the platform took action on their reports. "It's not made transparent to the user at all what exactly happens when you file a report... It's not clear to what extent Meta is using machines or humans," said Callum Hood, the head of research at the CCDH. Regardless of the mechanism used around reports, he says it's not working, and more transparency is necessary. Meta did not immediately respond to Mashable's request for comment. The CCDH criticizes Instagram's failure to act as it normalizes abuse and leads to repeat offenders -- one in five of the 1,000 worst comments were posted by repeat offenders -- and urges lawmakers to take action to hold Instagram accountable. Ahmed reiterated that there is no place for hateful abuse in political discourse, and as social media is increasingly central to that discourse, we should increasingly pressure social media companies to act.