Wednesday, 31 October 2018

Computational propaganda, bots, and the amplification of anti-semitism online

Just days before the horrific mass murder at Pittsburgh's Tree of Life synagogue, my Institute for the Future colleagues Sam Woolley and Katie Joseff published a deeply upsetting study on how social media bots and computational propaganda are being used to instigate and amplify anti-semitism online and manipulate public opinion. From the paper:

This report explores the ways in which online propaganda, harassment and political manipulation are affecting Jewish People in the runup to 2018 U.S. midterm elections. In the course of our research, members of this group have described a marked rise in the number of online attacks their community is experiencing. This is proving especially true during electoral contests and major political events. Correspondingly, our analyses suggests that tools like social media bots, and tactics including doxxing, disinformation, and politically-motivated threats, have been used online during the 2018 midterms to target Jewish Americans. According to interviewees, veiled human users—rather than automated accounts—often deliver the most worrisome and harmful anti-Semitic attacks.

As part of the wider paper series focused on “humanizing the effects of computational propaganda” this empirical work details the ways in which the Jewish socio-religious population in the U.S. is being disproportionately targeted with disinformation and abuse during this crucial political moment. We use a mixed methods approach in this research, deploying both qualitative and quantitative analysis in order to generate both a culturally deep and statistically broad understanding of how computational propaganda is being leveraged against this community...

Analysis of 7,512,594 tweets over a period from August 31, 2018 to September 17, 2018 shows the prevalence of political bots in these efforts and highlights groups within the U.S. political spectrum most involved in anti-Semitic attacks. In fact, as many as 30 percent of the accounts messaging using derogatory terms gathered in this data set appear to be highly automated.

"Computational Propaganda, Jewish-Americans and the 2018 Midterms: The Amplification of Anti-Semitic Harassment Online" (ADL)

Learn more at IFTF's Digital Intelligence Lab.