The rapid expansion of Artificial Intelligence (AI) and Data Science is transforming criminology by introducing predictive models capable of supporting decision-making in policing, criminal justice, and public security. This study critically examines the growing use of predictive analytics, machine learning, and algorithmic risk assessment systems in criminological practice, highlighting both their potential and their limitations. Through the analysis of contemporary literature, regulatory frameworks, and international case studies, the paper explores how data-driven approaches are reshaping crime prevention, recidivism prediction, resource allocation, and judicial decision-making. Predictive systems can process large and heterogeneous datasets, identify hidden patterns, and generate insights that enhance operational efficiency and evidence-based policymaking. However, these technologies also raise significant concerns regarding algorithmic bias, transparency, accountability, privacy, and the protection of fundamental rights. Historical datasets often reproduce structural inequalities, causing predictive models to reinforce discriminatory practices under the appearance of technological neutrality. Furthermore, the increasing adoption of complex “black-box” algorithms challenges explainability and due process, particularly when automated predictions influence decisions that affect individual liberty. The paper argues that the future of criminology depends on balancing technological innovation with ethical governance, democratic oversight, and human rights protections. Emerging approaches such as Explainable AI, fairness-aware machine learning, algorithmic auditing, and Human Rights Impact Assessments are presented as essential mechanisms for ensuring responsible deployment. Ultimately, AI and Data Science should function as decision-support tools rather than substitutes for human judgment, contributing to a more effective, transparent, and socially responsible criminal justice system.

Keywords

Artificial Intelligence, Data Science, Predictive Criminology, Algorithmic Decision-Making, Criminal Justice.