Racial algorithmic discrimination and human rights in the use of artificial intelligence

impacts of PL 2.338/2023

Abstract

Racial prejudice is still present in the Brazilian reality, unfortunately, and because of this discriminatory conduct the programming of algorithms used by Artificial Intelligences suffers a negative influence in some cases. The seriousness of the current situation of the theme in the country was the subject of debate in an event held in May 2024 by the Information and Coordination Center of Ponto BR - NIC.br, an agency linked to the Internet Steering Committee in Brazil. The research reported here has collected information and analyzed recent documents on attempts to offer greater security and respect to people on social networks, especially more vulnerable groups. It describes and evaluates Bill 2.338/2023, which proposes a Legal Framework on Artificial Intelligence for the country, in line with what has already been done by the European Union, the first in the world to create a standard on the subject. Therefore, faced with the problem of lack of proper regulation, discriminatory offenses and biases may continue to occur. The objective of the article is to analyze the potential impact of the proposed legislation, in parallel to the need for respect and inclusion defended by the Federal Constitution of 1988 and the 2030 Agenda. The theoretical framework is the principle of human dignity, as explained by Ana Paula Barcellos, and Tarcízio Silva’s analysis of racism algorithm. The dialectical method (perceived-conceived-lived) was used to develop the research and the technique of comparative study between legislations, in addition to the documentary review. It is concluded that there is a regulatory gap on algorithmic racism in Brazil, damaging the human dignity of these vulnerable groups.