THE PROFILE OF STUDENTS' SELF-EFFICACY IN CHEMISTRY LEARNING DURING THE TRANSFORMATION OF GENERATIVE AI USAGE

Abstract

The integration of Generative Artificial Intelligence (GAI) in chemistry education offers an innovative shift in teaching abstract concepts and high cognitive demands. This study aims to investigate the self-efficacy profile of students related to the application of GAI in chemistry learning. Using a descriptive quantitative approach with a survey design, this study involved 44 students of the Chemistry Education Study Program at the Islamic University of Indonesia, Yogyakarta, who completed a validated self-efficacy questionnaire. This instrument measures three dimensions of self-efficacy, namely level, generality, and strengthData analysis shows that most respondents (38.64%) have self-efficacy in the medium category with an average score of 67.68, while 29.55% are in the low category (63.94). As many as 20.45% of respondents have high self-efficacy with an average score of 72.76. The dimension of optimism towards AI-assisted task completion showed the highest score (70.55), reflecting students' positive perception towards AI as a learning partner. However, the predominance of moderate and low self-efficacy levels indicates challenges related to digital literacy, technology anxiety, and suboptimal pedagogical integration of GAI. These findings indicate that the effectiveness of GAI in enhancing academic confidence and self-directed learning is contingent upon adequate instructional support, effective faculty facilitation, and institutional preparedness. Thus, GAI implementation in education must prioritize the development of a pedagogical framework that considers students' psychological readiness and digital competencies