To Guide or Regulate? An Analysis of Brazilian Public University Documents on Academic Generative AI Use

Abstract

In recent years, research and development in Artificial Intelligence (AI) have taken on new configurations, especially after the launch of ChatGPT in October 2022. In the context of universities and research institutes, there is growing concern about the mediation of AI and issues related to ethics, authorship, plagiarism, costs, and the dissemination of research results. In this scenario, this article aims to analyze official documents, such as ordinances, resolutions, and guidance documents produced by public Higher Education Institutions in Brazil, seeking to understand how aspects related to transparency, data protection, authorship, rights, and biases are addressed in the documents produced by the 27 institutions analyzed. The research question guiding this study is: how have Brazilian public universities guided and/or regulated the use of generative artificial intelligence in the academic context? Based on a qualitative, bibliographic, and exploratory approach, the analysis indicated that transparency constitutes a central requirement in the documents, as does the concern with biases, understood as dimensions related to the mitigation of prejudice and algorithmic discrimination. It is concluded that the discussion on the use of generative artificial intelligence in Brazilian institutions is still occurring timidly, and that the existing documents are characterized by two main dimensions: one normative-regulatory and the other guidance-formative