Reconhecimento facial em tempo real e segurança púbica: perspectivas a partir do Regulamento Europeu para construção de modelo condizente com direitos humanos

Abstract

The present work aimed to evaluate the compatibility of the European Artificial Intelligence Act regarding the use of facial recognition technology for public security purposes with the protection of Human Rights and its eventual compatibility with the Brazilian legal system, in order to serve as a source of legislative inspiration. Based on the perspective that public security activities represent multigenerational fundamental and human rights, from which the obligation of state action arises, it was sought to determine how facial recognition technology, driven by artificial intelligence, could assist in such a task. It was pointed out that facial recognition encompasses several technologies aimed at different uses and purposes. The present work focused on remote biometric identification in real-time, which consists in the instantaneous identification of an individual in a crowd. The controversial technical and legal aspects of this technology, which follows the life cycle steps of artificial intelligence, were dissected. Having established the premise, based on the theoretical framework, that it is possible to regulate such technology for restricted uses, the European Artificial Intelligence Act was approached, first contextualizing the diploma and then going into specific rules about the object of study. The relevance of the European Artificial Intelligence Act arises from the depth of the normative diploma, the brazilian tradition in endorsing European legal thought and the so-called Brussels effect. Regarding remote biometric identification in real time, the aforementioned Regulation prohibited its use for public security as a general rule, except in exceptional cases, which also depend on compliance with strict guidelines and strict procedures before a judicial or similar authority for approval. It was considered that the Regulation, despite possible points of improvement, appears, in general, to be attentive to human and fundamental rights, without prejudice to potential betterments throughout its implementation. In Brazil, although the technology is already being used, there is a lack of regulation. In view of the European proposals, it was concluded that, although the Regulation is a relevant normative source, its generic provision at certain points, without addressing specific issues of real-time remote biometric identification and its particular risks, ended up leaving loopholes in the implementation of the system that could lead to violations of fundamental rights. Thus, in addition to the provisions of the European Regulation, suggestions were proposed to strengthen regulation in the brazilian national scenario in favor of protecting the fundamental and human rights involved