THE USE OF ARTIFICIAL INTELLIGENCE IN URBAN MONITORING: main challenges to the protection of legal assets and privacy

PRINCIPAIS DESAFIOS À TUTELA DOS BENS JURÍDICOS E À PRIVACIDADE

Abstract

ABSTRACT: Artificial intelligence (AI) represents a scientific vanguard with the potential to redefine various social sectors. In the realm of public security, its application has gained traction in several Brazilian states, aiming to enhance urban monitoring. However, the implementation of these technologies, especially through so-called "public security totems," has presented significant material errors. Such inaccuracies have posed challenges to the protection of legal assets and citizens' privacy. For example, imprecise facial recognition can lead to the erroneous identification of innocent people, resulting in embarrassment, unjust detentions, and even irreversible damages. Meanwhile, criminals remain unpunished, compromising the effectiveness of security that may exhibit flaws.

Given this truly concerning scenario, the investigation analyzes the increasing use of AI in urban monitoring and the possible impacts generated by its failures. In this sense, the problem arises: in cases of urban monitoring used with the aim of expanding the effectiveness of public security, how can the inaccuracies of artificial intelligence compromise the protection of legal assets concerning privacy and human dignity? The central hypothesis is based on the application of AI in urban monitoring systems which, in cases of lacking technological refinement and specific, adequate legal regulation, compromises the respect and observance of constitutional principles and individual rights, resulting in multiple privacy failures. The general objective is to analyze the main challenges that have arisen due to the use of AI in urban monitoring in Brazil, especially concerning the protection of legal assets, privacy, and public security, proposing reflections on the improvement of legislation inherent to the technology.

The methodology used is based on the inductive method, aiming to observe specific cases of urban monitoring that result in illegal constraint and violations of fundamental rights, to present broad conclusions about the need for technological and legal adaptation. Regarding the subject, the bibliographic survey covers doctrines on AI, public security, and constitutional rights, eminently aiming to build a theoretical basis that allows a qualitative analysis of the researched content. To this end, works such as those by Machado (2023), which addresses Law and AI; CNJ reports (2024) on the use of Generative AI in the Judiciary; analyses from ISCBrasil (2024) on AI as an ally of public security; articles by Freitas (2025) and Américo (2024) that discuss "security totems" and their impacts; case reports such as that documented by Data Privacy Brasil (2025) on erroneous identifications; and the Federal Constitution of 1988 itself, essential for research on fundamental rights and guarantees, were consulted. Finally, the data analysis and resulting discussions lead to considerations of perspectives in which a reformulation in the employment of these technological means is imperative, aiming at a more equitable, respectful, and less embarrassing approach, prioritizing constitutional principles and privacy, simultaneously with the improvement of public security effectiveness.