Social media, extremism, and radicalization
Published By admin with Comments 0
In the U.S., the sentencing of the 2018 Pittsburgh Tree of Life synagogue shooter and the arraignment of the former president for his role in the conspiracy and riots of 6 January 2021 dominated the news in early August 2023. As both incidents were partly inspired, planned, and documented in extremist networks on social media, public reflections about such events should rekindle questions about the role of social media in extremist radicalization in American public life.
Many would blame social media platforms—in particular, their algorithms that sort and recommend content—for the spread of extremist ideas. However, empirical evidence, including a study of YouTube led by Annie Y. Chen in this issue of Science Advances, reveals a more complex reality (1). The platforms and their algorithms rarely recommend extremist content, yet they remain powerful tools for those who hold extremist beliefs. Radicalized users can still use social media to access and disseminate ideas, build solidarity, or plan and publicize egregious acts. Indeed, despite efforts to remove or reduce the visibility of extremist content, social media platforms like YouTube continue to provide a hospitable environment for content espousing violence, hate, and conspiracist thinking of various kinds (Fig. 1).
Critical accounts of the ills wrought by social media have become commonplace, but the details are still important. YouTube, launched in 2005 and acquired by Google (now Alphabet) in 2006, is one of the most popular social media platforms in the United States (2). YouTube’s recommendation algorithms, which drive massive amounts of content consumption on the site, have a notorious reputation for surfacing hate speech, unfounded rumors, misinformation, hoaxes, and conspiracies. The platform’s recommendations, so the story goes, expose casual users to extremist content, nudging them down “rabbit holes” of (usually right wing) radicalization.
Content retrieved from: https://www.science.org/doi/10.1126/sciadv.adk2031.
