O marco regulatório da inteligência artificial no Brasil e o modelo de autoridade competente para implementar e fiscalizar a lei

Abstract

In the context of a hypercomplex global society, controversial questions arise about Artificial Intelligence (AI), an emerging technology with the potential to significantly transform society. Discussing AI, however, is not something new. However, in the context of Covid-19, several changes occurred throughout the world dynamics that accelerated digital transformation and technological development. This acceleration has had repercussions on the exponential evolution of AI in recent years. Thus, several dilemmas that include futurists, were the subject of reflections in the field of Law, among them, issues involving threats to jobs, use for health issues, autonomous cars, facial recognition, privacy and protection of personal data, the impacts on the education and society broadly and, especially, the regulation of AI. With this, the main objective is to examine Brazil's position in the AI regulatory situation and what is the competent authority model to implement and supervise the regulatory framework for this technology in Brazilian territory. Thus, to enable the development of the thesis, initially, this work aims to analyze whether there is a universally valid concept about AI and, to this end, it uses a transversal approach focusing on the technological, ethical, economic and legal axes, considered the main ones for this thematic section. Next, regulatory initiatives and AI authorities at global, regional and local levels are analyzed. It focused, therefore, on the initiatives of the UN, UNESCO OECD, EU, USA and China, as a way of exemplifying existing regulations and some dilemmas arising from normative initiatives. Finally, the Brazilian regulatory situation on AI was examined, in which, at present, PL nº 2338/2023 represents the regulatory framework for technology and an overview of the topic was offered, to later highlight predictions about the authority AI authority and discussions about the need to create this authority or whether any previously existing State structure will be used, such as the ANPD, for example. It is concluded that the ANPD should not assume the position of competent AI authority and that the specificities of this technology exceed the capacity of the agencies currently existing in Brazil.As an alternative, the preliminary institutional design of the competent authority was shown to be adequate to the complexity of AI technology. To make this possible, a methodology was adopted that covers a bibliographical study, in books, articles, dissertations, theses, various productions, videos, AI platforms, in addition to the method developed by Nitish Monebhurrun for the formulation of legal texts and sought - applying Simple Language and Visual Law techniques in the construction of the thesis to make the topic more accessible, based on the notion that Law is for everyone.