Discriminação algorítmica e o direito à explicação na Lei Geral de Proteção de Dados Pessoais

Abstract

The current scenario of development of information and communication technologies has spread the use of mobile electronic devices that are constantly collecting data and distributing the personal data of their holders on the internet. With this, an ocean of data was formed that can be collected, organized into categories or profiles, processed and finally new information is created based on this primary data. Therefore, we may receive more personalized services or products that turn the wheel of the shared economy. However, in addition to the benefits listed here, the possible risks of misuse of personal data do not go unnoticed. Considering that among the technologies for this treatment are artificial intelligence algorithms, Big Data and profiling techniques, new situations arise for Brazilian law related to the defense of fundamental rights and technological innovation. This work aims to understand how AI algorithms and profiling techniques can discriminate through automated decisions and how Law No. 13.709/2018 (General Law for the Protection of Personal Data) in its article 20 can defend data subjects by guaranteeing the right to explanation and review of exclusively automated decisions that affect the interests of these holders. For this, we sought to compile the factors that led to industry 4.0 and how the legislation of the European Union evolved towards the creation of the General Data Protection Regulation. Finally, it is verified how the Brazilian legal framework has some data protection institutes prior to the validity of the LGPD. Finally, this work analyzes some examples of discrimination committed by algorithms and demonstrates how the LGPD institutes defending the right to explanation and review of exclusively automated decisions as tools for combating algorithmic discrimination. In order to achieve the claim listed herein, basic research methodology and exploratory compilation with a qualitative approach were used, carrying out documentary research (articles of law and other normative acts). Bibliographic research and literature review (articles, doctrines and specialized journals) were also carried out. Keywords: automated decisions; algorithmic discrimination; artificial intelligence; data protection; profiling.