Legal Studies in Digital Age In Press 2027 07 01 Artificial Intelligence and the System of Responses to Crime: From Situational Prevention to the Engineering of Multilayered Sanctions Artificial Intelligence and the System of Responses to Crime: From Situational Prevention to the Engineering of Multilayered Sanctions 1 17 EN Journal Article 2026 07 21 <p>This article examines the role of artificial intelligence in the system of responses to crime, focusing on the continuum from situational and person-oriented prevention to the design of multilayered sanctions. The study proceeds from the distinction between the cognitive and analytical capacities of artificial intelligence and the normative authority required for legally significant decisions. In situational prevention, artificial intelligence can support dynamic analysis of spatial, temporal, and environmental risk patterns, improve resource allocation, and facilitate continuous evaluation of preventive interventions. However, predictive outputs must remain distinct from automatic intervention because crime data may reflect previous patterns of surveillance, reporting, and enforcement and may generate self-reinforcing feedback loops. In person-oriented prevention, the legitimate function of artificial intelligence lies not in identifying presumed future offenders, but in analyzing risk factors, protective factors, vulnerabilities, and modifiable needs in order to support proportionate and non-stigmatizing interventions. The article further examines the transition from prevention to reactive criminal policy and argues that artificial intelligence may assist in comparing criminalization, administrative regulation, and non-punitive alternatives without independently determining the legal status of conduct. In the field of sanctions, artificial intelligence can contribute to the engineering of multilayered responses by comparing criminal, administrative, disciplinary, professional, rehabilitative, medical, psychological, and social measures and by identifying less intrusive alternatives where appropriate. The central argument is that artificial intelligence should operate as a bounded decision-support infrastructure rather than an autonomous decision-maker. Its legitimate value lies in strengthening evidence-based analysis, proportionality assessment, monitoring, and policy revision while preserving human oversight, legal responsibility, contestability, and fundamental rights. The article ultimately proposes a cyclical model of identification, analysis, intervention design, implementation, evaluation, and revision for the responsible integration of artificial intelligence into criminal policy.</p> https://www.jlsda.com/index.php/lsda/article/download/526/447