Direito a inferências razoáveis como substrato normativo na consolidação de mecanismos de governança algorítmica nos sistemas automatizados de decisão

• Ano de defesa: 2022
• Autor(a) principal: Miranda, Paulo Rodrigo de
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Direito; UFSM; Programa de Pós-Graduação em Direito; Centro de Ciências Sociais e Humanas
• Palavras-chave em Português: Direito a inferências razoáveis; LGPD; Sistemas automatizados de decisão; Transparência; Accountability; Right to reasonable inferences; Automated decision systems; Transparency; CNPQ::CIENCIAS SOCIAIS APLICADAS::DIREITO
• Link de acesso: http://repositorio.ufsm.br/handle/1/24915

This study sought to verify the feasibility of recognizing the right to reasonable inferences as a normative substrate in developing algorithm governance mechanisms in automated decision-making systems that seek to strengthen ex-ante approaches and a normative structure permeated by a risk-based approach that ensures greater transparency and accountability on the part of data controllers. Four chapters containing the specific objectives were outlined to develop the dissertation’s theme. The first chapter addresses issues involving the society of control in the digital age within the context of algorithmic governmentality to encourage reflection between the power of algorithm control in automated decision-making systems and the freedom of choice of the recipients of these systems, as well as an analysis of the risks arising from implementing these systems, including discriminatory biases and opacity. The second chapter covers the constitutional elements of the protection of personal data within a proposition of amplifying its scope of protection to recognize the right to reasonable inferences as an unfolding of the subjective dimension of the fundamental right to data protection by analyzing the subjective and objective dimensions of the referred fundamental right. The third chapter presents the right to reasonable inferences as a normative substrate in consolidating governance mechanisms. Lastly, the fourth chapter encompasses an analysis of the role of algorithm governance within a framework of co-regulation and a risk-based approach, which seeks to consolidate the institutional arrangements of the General Personal Data Protection Law in establishing compliance and accountability guidelines by data controllers involving automated decision-making systems. To answer the question of the topic, a deductive approach was employed with a generalization of the topic for the resumption of a particularized question; likewise, the bibliographical research method was used. It was possible to conclude that, summarily, the right to reasonable inferences can be understood as a standard of behavior for the data controller and having the capacity for establishing governance mechanisms based on compliance. This will enable the development or implementation of automated decision-making systems to be assured and employed by adopting technical and organizational measures that are appropriate to the risks involved in their activities, enabling, as much as possible, transparency and accountability in this proceduralization.