Regulação de sistemas de inteligência artificial: papel do Estado no ambiente regulatório a partir da modernidade periférica

Abstract

The various systems covered by the broad concept of Artificial Intelligence (AI) permeate practically all aspects of social life today, from the most common applications to highly complex tasks, providing numerous social and economic benefits. However, these same systems present challenges in that they are capable of violating individual and collective rights that are fundamental to a Democratic Rule of Law. Therefore, it is necessary to question the construction of appropriate regulatory mechanisms by the State, capable of enhancing the benefits and reducing the risks associated with the development of these technologies. More than that, it is necessary to analyze the aspects that permeate and enable the adequate regulatory capacity of these mechanisms. Given this, the central and guiding question of this work is: how to build appropriate regulatory strategies for the Brazilian State that best meet the technical specificities of Artificial Intelligence (AI) systems, considering the socioeconomic context and the structural problems faced by Brazil? To answer this question, the work is based on the contributions of the Economic Analysis of Law and the sociological perspectives of the economy initially, to understand the foundations of regulation and how it develops in the Brazilian context. Furthermore, the idea of Peripheral Modernity, constructed by Neves based on Systems Theory, is adopted as a theoretical background framework, and how this interferes with the capacity for regulation. Next, the role of different social actors in regulation and technological development is discussed, as well as possible institutional arrangements aimed at encouraging innovation and realizing rights. Finally, based on this theoretical framework, Brazilian regulatory initiatives are critically analyzed, both by the Legislative Branch and the Executive Branch, to verify the strategies and mechanisms that allow the development of a regulatory environment capable of dealing with the uncertain and changeable character inherent to regulations and technical peculiarities related to AI systems. It is concluded that although a normative-legal framework is fundamental to protect individuals in the face of advances in artificial intelligence, the effectiveness of these protections transcends the simple promulgation of laws. It is therefore necessary to take broader and more strategic action to deal with regulatory problems and incentives for internal development.