Tracking changes in pharmacy students' engagement with ChatGPT: A comparative cross-sectional study (2023-2024).

INTRODUCTION: Since its launch, ChatGPT has generated both interest and concern in education. Between 2023 and 2024, institutional policies shifted from restrictive to conditional use, increasing student exposure. This cross-sectional study compared independent cohorts from those years to assess differences in knowledge of AI and in impressions and appropriate use literacy regarding ChatGPT, and to examine motives for use, reasons for not using, concerns, and perceived need for AI education. METHODS: Online surveys were conducted in 2023 and 2024. The questionnaire assessed knowledge of AI, impressions and appropriate use literacy regarding ChatGPT, with acceptable internal consistency (Cronbach's α > 0.70). Most items were retained for comparison, with 2024 updates reflecting OpenAI policy changes. Students were grouped by ChatGPT awareness (able to describe ChatGPT vs not). Fisher's exact test and the Mann-Whitney U test were applied; qualitative responses underwent thematic analysis. RESULTS: A total of 113 responses were analyzed in 2023 and 82 in 2024. Students who could describe ChatGPT demonstrated higher knowledge of AI, whereas appropriate use literacy did not improve across cohorts. In 2024, 62.2 % supported integrating AI education. Reported concerns included reliability, overreliance, lack of guidance, and privacy. Non-users cited lack of necessity and technical barriers. Thematic analysis highlighted both benefits and challenges of ChatGPT use. CONCLUSIONS: Despite increased exposure, ChatGPT use and AI literacy remain limited. Structured AI education is essential to promote responsible and effective use in pharmacy education.