Vigilância algorítmica e discriminação abusiva: uma análise do Sistema SMART SAMPA à luz do princípio da não discriminação da LGPD (art. 6º, IX) 278 Tais erros, decorrentes da margem de incerteza da tecnologia, não são meras falhas técnicas, mas sim uma violação de direitos fundamentais como a liberdade e a dignidade, caracterizando discriminação abusiva por efeito e responsabilizando objetivamente o Poder Público. PALAVRAS-CHAVE: reconhecimento facial; discriminação algorítmica; proteção de dados; SMART SAMPA; direitos fundamentais. ABSTRACT: This article analyzes the SMART SAMPA program, from the Municipality of São Paulo, as its object of study, focusing on the inconsistencies of its facial recognition system. The central problem is the configuration of abusive discrimination through false positives, which violate the principle of non-discrimination of the General Data Protection Law (LGPD). The hypothesis is that operational flaws, especially identification errors, lead to disproportionate and harmful data processing that infringes on fundamental rights. The general objective is to analyze how these inconsistencies, notably false positives, characterize abusive discriminatory treatment of sensitive personal data. The specific objectives include: contextualizing the evolution of surveillance; defining the applicable LGPD principles; identifying the flaws of the SMART SAMPA based on its transparency report (Nov 2024 – May 2025); and demonstrating how such flaws constitute algorithmic discrimination. The adopted methodology was legal-dogmatic and qualitative, with normative analysis, bibliographic review, and documentary analysis of the program’s report. The research findings indicated that, of the incidents generated by inconsistencies, 28.05% (23 cases) were false positives from facial recognition, resulting in unlawful police detentions. Such errors, arising from the technology’s margin of uncertainty, are not mere technical flaws but a violation of fundamental rights such as liberty and dignity, characterizing abusive discrimination by effect and establishing the objective liability of the Public Administration. KEYWORDS: facial recognition; algorithmic discrimination; data protection; SMART SAMPA; fundamental rights. INTRODUÇÃO No âmbito das discussões contemporâneas sobre a intersecção entre inteligência artificial e a proteção de direitos, este artigo
RkJQdWJsaXNoZXIy MjEzNzYz