Dark Cycles: Social Engineering and Political Chatbots in Netanyahu’s 2019 Election Campaigns

This study investigates the potential adverse effects of political chatbots operating as peer-to-peer propaganda tools that evade public and regulatory oversight. We analyze a specific case of a Facebook Messenger chatbot used by Benjamin Netanyahu’s campaign during 2 Israeli elections in 2019. Applying the Walkthrough Method, we define “dark cycles” as a convergence of social engineering and dark patterns, characterized by 3 phases: reconnaissance (establishing connections and collecting data), training (using repetitive messaging and dark-pattern choice architectures to shape perspectives), and activation (instructing users to perform specific tasks). Although centered on this case, our findings suggest broader implications for studying political chatbots as AI technologies evolve. We also address the asymmetric power dynamics these chatbots create, highlighting their role in shaping compliant data subjects under surveillance and automation.