According to a study released on Friday by the Center for Countering Digital Hate, hate speech against Black people, transgender people, and engagement with hateful content on Twitter have all drastically increased since Elon Musk took control of the platform. The study’s findings are now being considered in the context of Twitter’s growing dependence on automated moderation rather than human involvement to tamp down on hate speech on social media platforms.
The results contradict Musk’s assertions from last week that “hate speech impressions” on Twitter have decreased by a third since he took over the company’s leadership. Musk, in response to a New York Times article on the study, tweeted on Saturday that the results are “utterly bogus.”
Engagement with offensive content increased by 372%
Hateful tweets will be “max deboosted & demonetized,” Elon Musk tweeted in mid-November. As a result, Twitter users won’t see the content unless they search for it on the site.
But the study discovered that after Musk took over the company, interaction with hate speech skyrocketed to indiscriminate levels. In the days before Musk took control of the network, there were 13.3 likes, responses, and retweets on average for tweets that contained offensive language.
According to the analysis, average engagements on nasty content have increased dramatically since Musk took control, by a whopping 372%, to 49.5.
An unsettling hate speech against Black people
A daily average of 1,282 messages featuring epithets against Black people were posted on Twitter before Elon Musk took control of the service. This figure increased to 3,876 tweets after Musk officially acquired Twitter. The number of tweets using racial insults against Black people increased even further to 4,650 tweets last week, according to the data, despite Musk’s assertion that hate speech impressions had decreased by a third.
Hate speech toward transgender people has increased by 62%.
The study discovered that since Musk acquired control of Twitter, slurs directed at transgender people—with 5,117 posts on average—have surged by 62%. The information comprised tweets in English from all over the world and was acquired using Brandwatch, a social media analytics platform
Why exactly is Hate speech surging on Twitter?
According to Ella Irwin, vice president of trust and safety products at Twitter, the network is eliminating some manual reviews rather than completely banning specific speech. Now, the platform mainly relies on automation to control content.
She claimed that Musk was concentrated on increasing automation and argued that the business had previously erred by relying too heavily on labor- and time-intensive human assessments of damaging content.
“He has encouraged the team to take more risks, move fast, get the platform safe,” Irwin said.
Despite ongoing worries from civil rights organizations about how hateful content could proliferate on Twitter, Elon Musk’s act of erratic firings of the platform’s existing staff (by over 50%) since taking over Twitter in late October has reportedly destroyed the trust and safety teams in charge of moderation.