Abstract
The increasing use of artificial intelligence (AI) in policing encompasses predictive analytics, surveillance systems, facial recognition, and other data-driven decision-support tools. While these technologies might facilitate operational efficiency and enhance crime-prevention capabilities, they also raise serious ethical and accountability issues. This study adopted a qualitative secondary desk study, using a structured, systematic literature review to examine the intersection of AI, ethics, and accountability in policing. The review synthesized evidence from peer-reviewed journal articles, institutional reports, and governance frameworks published between 2015–2025. The findings are organized into four major thematic areas: bias and discrimination in algorithms, transparency and explainability issues, privacy and surveillance risks, and gaps in accountability and oversight. The results indicated that policing AI exhibits the capacity to reproduce historical bias inherent in training data, serve as black boxes for decision-making processes and widen surveillance capabilities in ways that meaningfully threaten privacy and civil liberties. The study also found a chronic divide between high-level ethical values and their operationalization in policing institutions. This paper also contributed to the existing body of knowledge by integrating ethics and governance into a single analytic framework, before outlining implications for police legitimacy, institutional readiness, and responsible AI adoption, particularly in developing contexts.