Artificial intelligence (AI) has shown a unique ability to tailor and enhance personalized learning experiences. However, its application in physical education is limited. This study aims to investigate the effect of the AI-driven application on the learning outcome and student engagement among physical education students in higher education. A quasi-experimental two-group design was used to determine differences between pre-test and post-test scores. Standardized tests that measured musculoskeletal strength and endurance were utilized. The Utrecht Work Engagement Scale for Students (UWES-9S) assessed student engagement. For eight weeks, the experimental group (EG) had an AI-driven physical activities while the control group (CG) followed the usual PE class. Post-test scores showed the experimental group significantly improved the learning outcome (p<0.001). Engagement scores were also significantly higher in EG (p < 0.01). Consequently, The CG showed a significant decrease in student engagement while the increase in learning outcome is not statistically significant. These findings suggest that AI-driven applications positively affect students’ physical fitness and engagement. Further research is recommended to fully implement AI in classrooms, addressing student and teacher literacy and ethical concerns to maximize its potential in the PE curriculum.

Keywords: Artificial Intelligence (AI), physical fitness, student engagement, learning outcome, tertiary education