Regulating Predictive Policing: Balancing Public Security and Civil Liberties in Algorithmic Law Enforcement

Abstract

Predictive policing has emerged as one of the most influential—and contested—applications of algorithmic decision-making in contemporary law enforcement. This narrative review examines the conceptual foundations, regulatory challenges, and governance implications of predictive policing as states increasingly rely on data-driven systems to forecast crime and allocate enforcement resources. The analysis synthesizes interdisciplinary scholarship across criminal law, sociology, digital governance, and public-security policy to illuminate how predictive policing transforms traditional policing practices and introduces new forms of algorithmic power. The review demonstrates that while predictive systems promise operational efficiency and targeted interventions, their effectiveness in reducing crime remains mixed, often shaped by short-term gains, displacement effects, and feedback loops created by historically biased data. Beyond questions of accuracy, predictive policing raises significant concerns associated with discrimination, opacity, due process, privacy, and the expansion of surveillance infrastructures. Comparative examination reveals that regulatory responses vary widely across jurisdictions: the United States exhibits fragmented municipal and federal approaches; the European Union has adopted rights-centered, high-risk regulatory models; the United Kingdom and Commonwealth rely heavily on judicial oversight; and regions in the Global South face unique democratic and institutional vulnerabilities. Drawing from these observations, the review highlights the need for robust governance frameworks grounded in transparency, necessity, proportionality, contestability, and public accountability. Ultimately, the study argues that predictive policing must be regulated through principles that safeguard civil liberties while enabling legitimate public-security objectives, emphasizing that the long-term legitimacy of algorithmic law enforcement depends on democratic oversight, institutional readiness, and meaningful community participation.