AI AND ENGAGEMENT DETECTION: ATTENTION AND PARTICIPATION (BENEFITS AND RISKS)

Abstract

 This study investigated the use of artificial intelligence to detect student engagement in digital educational environments, emphasizing attention and participation and considering benefits and risks. The problem was defined as understanding to what extent such use contributed to pedagogical practice without reinforcing surveillance, assessment distortions, and ethical risks related to monitoring. The general objective was to analyze the benefits and risks of using artificial intelligence for engagement detection, discussing ethical and pedagogical limits to prevent the consolidation of surveillance practices in digital educational contexts. A bibliographic research methodology was adopted, with qualitative analysis of selected academic literature on education in the digital era, technologies, online assessment, and teacher education. In the development, the complexity of engagement and the inadequacy of reducing it to operational metrics were discussed, as well as the possibility of using indicators as support for teaching mediation when interpreted contextually and for formative purposes. Risks related to surveillance, privacy, bias, and the reconfiguration of assessment practices were also problematized. In the final considerations, it was concluded that AI-based engagement detection was potentially beneficial when used as complementary evidence to pedagogical judgment, with clear limits and transparency, but tended to reinforce control and assessment distortions when applied as an automatic mechanism for classification and evaluation, indicating the need for empirical studies to deepen the findings.