Learning with algorithms: Artificial Intelligence (AI) usage, trust, and literacy among university students

Abstract

As Artificial Intelligence (AI) becomes increasingly integrated into higher education, understanding the mechanisms behind students’ effective use of AI tools is critical. This study examines the impact of AI usage behavior on AI usage literacy, with critical thinking as a mediator and trust in AI as a moderator. A quantitative approach using Partial Least Squares Structural Equation Modeling (PLS-SEM) was applied to data collected from 296 university students in Vietnam. The findings confirm that AI usage behavior positively influences both critical thinking and AI usage literacy. Critical thinking significantly contributes to AI literacy and mediates the relationship between usage behavior and literacy. Furthermore, trust in AI strengthens the relationship between AI usage behavior and literacy, with a significant moderating effect. These results emphasize the importance of cognitive engagement and trust calibration in AI-supported learning. The study offers theoretical and practical contributions to the development of AI literacy in higher education and highlights the need for pedagogy that promotes critical thinking and responsible trust in AI systems. These findings urge higher education institutions to embed AI literacy and critical thinking in curricula and provide training for educators to foster students’ ethical, reflective, and effective AI engagement.