This study surveyed 118 students from 12 AI‑related courses at our university to examine how generative artificial intelligence (GenAI) usage relates to perceived learning experiences. Cluster analysis identified four user groups: high‑frequency users who reported many benefits, low‑frequency users with little reliance and few benefits, and two moderate‑use groups showing different benefit levels. Comparisons were made between free‑ and premium‑version users, single‑tool versus multi‑tool users, and students under varying instructor policies. Multivariate regression revealed that academic benefit was positively linked to early reliance and academic‑task support; positive impact correlated with cognitive reliance, task support, confidence in GenAI reliability, and supportive instructor policy; negative impact was associated with early reliance and attitudinal change. The adverse effect of early reliance intensified as evaluation literacy increased. Perceived benefits of GenAI‑enhanced learning centered on cognition, performance, and self‑efficacy, while concerns about stress and reduced critical thinking corresponded to lower perceived learning gains. The findings suggest institutions should craft clearer policies to mitigate inequities and enable students to harness increasingly capable AI systems.
Review