CRIME PREDICTIVE MODEL FOR DYNAMIC ALLOCATION OF RESOURCES IN THE PMPR: A LEGAL-TECHNOLOGICAL ANALYSIS ON EFFICIENCY AND PREVENTION OF ALGORITHMIC BIASES

Abstract

The contemporary era is marked by the inexorable digitization of social relations and the increasing incorporation of artificial intelligence-based technologies in Public Administration. In this scenario, public security emerges as a fertile ground for the application of innovations that promise to optimize resource management and enhance crime prevention. This scientific article addresses the complex and multifaceted issue of implementing predictive crime models within the Military Police of Paraná (PMPR), focusing on the dynamic allocation of ostensive policing resources. The research adopts a legal-technological approach to analyze the subject, investigating the dialectical tension between the constitutional principle of administrative efficiency and the imperative safeguarding of fundamental rights and guarantees. The main objective is to analyze the legal viability and ethical responsibility
of adopting such systems, proposing guidelines for their design and implementation in order to mitigate inherent risks, notably algorithmic biases and their discriminatory consequences. Through a methodology of bibliographic and documentary research,
with an analytical-deductive approach, the study explores the technological foundations of predictive policing, the challenges posed to personal data protection under the General Data Protection Law (LGPD), and the impacts on fundamental principles such as equality, presumption of innocence, and due process of law. The defended hypothesis is that the implementation of a predictive model in the PMPR is legally plausible and potentially beneficial; however, its legitimacy is strictly conditioned on the adoption of a robust governance framework, encompassing everything from the curation of training data to the implementation of mechanisms for transparency, auditability, and, crucially, the maintenance of the sovereignty of qualified human decision-making. The article concludes on the need for an interdisciplinary dialogue and careful regulation that balances technological innovation with democratic values, so that the pursuit of efficiency does not result in the automation of injustice and the deepening of historically consolidated social inequalities.