Human-Centered Security in Society 5.0: Revisiting Algorithmic Bias and Methodological Constraints in Predictive Policing

Abstract

Society 5.0, envisioned as an advanced, human-centered civilization in which the physical and digital realms seamlessly complement one another, represents a sustainable model for the functioning of human communities on an increasingly overpopulated planet. In such an environment, a progressively deeper symbiosis between social processes and technological innovation is expected, with the security sector emerging as one of the key domains of this integration. Within the field of public security, predictive policing stands out in particular: its models, based on artificial intelligence and big-data analytics, enable the anticipation of criminal patterns and the potential for more effective risk prevention. This paper examines predictive policing as a transformative approach to law enforcement, while simultaneously problematizing its deeply embedded challenges—especially in the context of the value framework of Society 5.0, which seeks to ensure that technological advancement remains subordinate to human dignity, justice, and social inclusion. Through an analysis of case studies from Chicago, London, and Tokyo, the paper identifies the operational advantages of predictive techniques but also highlights key concerns such as algorithmic bias, lack of transparency, insufficient data quality, and methodological limitations that may jeopardize the fairness and legitimacy of police interventions. The findings demonstrate that although predictive algorithms can contribute to the enhancement of preventive strategies, their implementation must occur within a clearly defined normative framework that incorporates technical robustness, independent oversight, institutional accountability, and active citizen participation. In line with the principles of Society 5.0, the paper concludes that the successful application of predictive policing requires the development of systems that are ethically grounded, methodologically transparent, and oriented toward the protection of human rights - ensuring that technology serves society rather than the other way around.