CCTV cameras are surveillance technologies that cities deploy to enhance urban safety. Increasingly, such systems are augmented with AI capabilities to facilitate advanced analytical and decision-making functionality. Research on CCTV systems has shown that these systems do not evenly ensure urban safety and that they may have adverse effects on marginalized populations. As risk profiling using CCTV systems gets augmented with AI, it becomes increasingly important to ask how these systems get built and how their design and implementation may enshrine such adverse effects. This presentation introduces India’s Safe City Project which aims to make cities safer for women through technological intervention. Focusing on the specific example of Bengaluru, a south Indian metro city where the Safe City Project is being implemented, we explore how values and power dynamics influence the design of AI-powered urban video surveillance systems. _x000D_
_x000D_
In this presentation, we discuss the preliminary findings from an ongoing research project. The project uses qualitative network interviews with different expert stakeholders to better our understanding of those power dynamics. Analytically, we focus on how incorporating AI into the CCTV system is discussed by expert stakeholders and how they describe the networks of people and objects involved in designing and implementing the system. We then build on those findings to develop a more nuanced understanding of the different values held by the actors involved in the making of the Bengaluru AI CCTV system. By doing so, we aim to identify hierarchies of whose values dominate the process and what kind of socio-technical power dynamics this entails. _x000D_

UDIPTA BORO; Fran Meissner; Karin Pfeffer
Department of Regional and Urban Planning and Geo-Information Management, Faculty ITC, University of Twente, The Netherlands
675