State and non-state political actors are using algorithms and automation in efforts to sway public opinion. In some circumstances, the ways coded automation interacts with or affects human users is unforeseeable. In others, individuals and organizations work to build software that purposefully targets voters, activists, and political opponents. Politicized social bots are one version of potentially malicious automated programs. Understanding how technologies like these are used to spread propaganda, engage with citizens, and influence political outcomes are pressing problems.
Recent findings confirm that there is bot activity in Germany: Angela Merkel was bombarded with bot-generated hate speech messages after the Berlin Christmas market attack; right-wing accounts on Twitter propagate xenophobic messages in relation to the German refugee debate; botnetworks supporting the AfD have been discovered on Facebook. What is more, after Merkel cautioned the German Bundestag in November 2016 against automated manipulation of opinion, there has been a heated discussion on how to regulate social bots. Parties and journalists are scrambling to come up with overblown proposals that might heavily restrict freedom of expression.
We have interviewed German bot developers, journalists, data scientists, policy makers, cyber-warfare specialists, and victims of bot attacks in order to investigate the application and potential impacts of computational propaganda in Germany; especially in relation to the Bundestagswahlen 2017, the ongoing refugee debate, right-wing populist politics and individual activism. We then work with computer scientists to detect and track bots on social media in Germany. With "real-time" social and information science we can examine automated propaganda in relation to events and debates, such as the Bundespraesidentenwahl 2017.