Invisible Eyes: Plausible Pathways for AI-Enabled Surveillance in North Korea – A Scenario Analysis

This paper examines how artificial intelligence could reshape surveillance and social control in North Korea, focusing on Pyongyang. Rather than viewing AI surveillance as a sudden rupture, it conceptualises it as a possible intensification of North Korea’s existing human-centred system, one that already includes household monitoring, workplace surveillance, mobility control, and ideological discipline. Drawing on studies of surveillance and algorithmic governance, and using the city of Urumqi in Xinjiang as an empirical reference architecture, the paper distinguishes three functions of AI-enabled surveillance: tracking, profiling, and pre-emptive policing. Using scenario analysis, the study identifies two critical uncertainties for the North Korean authorities: the level of domestic technological capacity and access to surveillance-relevant hardware, software, and expertise. It then constructs four potential scenarios for Pyongyang by 2035: the deployment of ‘symbolic’ AI, a constrained deployment of domestic AI, an import-dependent uptake, and a full AI panopticon. The paper argues that the most plausible near-term trajectory is a fragmented, import-dependent expansion of tracking and selective profiling rather than fully integrated predictive control. Despite their fragmentary nature, such systems could increase the costs for North Korean citizens engaging in unauthorised mobility, foreign media circulation, informal economic activity, and local non-compliance, while remaining constrained by energy shortages, fragmented data, bureaucratic silos, and everyday strategies of circumvention.