Abstract
The growing complexity of intelligence analysis, marked by high volumes of data and inherent uncertainty, has driven the integration of technology and statistics into crime prevention efforts. This article examines the evolution of predictive policing, with a specific focus on the Italian context and the Delia® software. Developed in Milan in 2007, Delia® combines statistical modeling, behavioral analysis, and geospatial data to forecast serial crimes. Unlike other systems, it requires human oversight for decision-making, thus balancing algorithmic precision with analyst discretion. The software has demonstrated measurable success, including a significant drop in robbery rates and improved case resolution in Milan. External evaluations, such as those conducted by economist Mastrobuoni, confirm Delia®’s effectiveness and highlight its cost-efficiency and potential to influence criminal behavior patterns. Despite these strengths, concerns remain regarding data integrity, false positives, and limited applicability to certain crime types. The article concludes by stressing the importance of a clear legal and ethical framework to guide the use of predictive technologies in law enforcement.