Social Control Through Data Science Systems: A Conceptual Essay

Boers, Klaus; Grimme, Christian; Huang, He; Kemme, Stefanie; Schaerff, Marcus; Singelnstein, Tobias

Abstract

Recent advances in data science, in particular, the analysis of big data through machine learning, have made it technically possible to realize concepts of total social control. However, social control must operate selectively to fulfill its functions of symbolization, integration, and reproduction which are crucial for the cohesion of a modern society. It is argued that selective social control is functional as it symbolically reinforces compliance with norms, prevents widespread stigmatization, and preserves resources and individual autonomy. Current trends toward the implementation of total control systems are illustrated regarding predictive policing, automated CCTV, and the Chinese social credit system. The technical potential and limitations of data science systems are discussed, as are the potential of the legal system to maintain resilience in the face of political and practical demands to expand control capacities. The authors highlight a creeping erosion of legislation and the possible normalization of comprehensive surveillance measures in the name of public safety. They conclude that only a selective and legally constrained use of data science can preserve the function of social control in open societies.