Mapping Current Use of Artificial Intelligence in Pharmacology Education via a Scoping Review

Pharmacology education, often reputed as complex, overtly didactic and decontextualised, may benefit from artificial intelligence-supported strategies. However, current guidance is fragmented across disciplines and contexts, thus weakening evidence-based curricular implementation. This scoping review mapped existing research to identify applications, strengths, limitations, and areas for future development. A double-blinded screening process facilitated by Covidence yielded 17 eligible studies from four databases. Studies mostly comprised cross-sectional studies from high-income countries in the medical context, with generative artificial intelligence being predominant (ChatGPT-3.5 and ChatGPT-4.0). Research comprised assessment (n = 12), paedagogy (n = 4), curriculum design (n = 1), and programme evaluation (n = 1). Most studies assessed tools' ability to answer examinations, with mixed success depending on the version, question type, and inclusion of context in prompt engineering. Few studies incorporated students, limiting insights into learning impact. Zero-shot prompting was mostly used, limited further by unclear design frameworks, which may bias outcomes considering downstream inefficiencies. Current research prioritises the performance of artificial intelligence, rather than its integration or impact in learning, which reduces its applicability for curriculum design and competency development. Although promising, the impact is limited, requiring clearer instructional design and rationalisation within the education ecosystem. Although there is considerable potential for pharmacology education, research requires greater structure, longitudinal design, and incorporation of students to inform clear impact. Purposeful, context-aligned implementation and continuous evaluation are needed to ensure ethical, valid, and meaningful use in pharmacology education. To support future research, recommendations are provided for practical reporting, scientific design, and impact measurement.