<![CDATA[The rapid adoption of ChatGPT in health-profession education has created new opportunities for personalized learning, clinical reasoning support, and academic productivity. However, concerns regarding information reliability, ethical use, and overreliance on artificial intelligence remain insufficiently understood, particularly in clinical pharmacy education. This study aimed to develop and validate a multidimensional instrument for assessing ChatGPT adoption among Indonesian pharmacy students and to examine the structural relationships among its determinants using Partial Least Squares Structural Equation Modeling (PLS-SEM). A cross-sectional study was conducted using convenience sampling between January – March 2026 among 398 pharmacy students from Indonesian higher education institutions with prior ChatGPT experience. The instrument comprised six domains: Knowledge, Perception, Perceived Usefulness, Concern, Application, and Ethics and Usage Boundaries. Content validity was excellent (I-CVI, S-CVI/Ave, and S-CVI/UA = 1.00), while internal consistency and convergent validity were satisfactory (Cronbach's α = 0.857–0.951, composite reliability (ρC) = 0.896–0.959, and AVE = 0.634–0.814). Structural model analysis revealed that Perception (β = 0.630, p < 0.001) and Perceived Usefulness (β = 0.367, p < 0.001) were positively associated with Concern, whereas Concern was negatively associated with Application (β = −0.201, p < 0.001). Perceived Usefulness was also negatively associated with Ethics and Usage Boundaries (β = −0.061, p = 0.021). These findings indicate that pharmacy students recognize both the educational value and potential risks of ChatGPT. The validated instrument may facilitate future evaluations of generative AI adoption and support responsible AI integration in pharmacy education.]]>
Copyrights © 2026
