INTRODUCTION: Artificial intelligence (AI)-generated educational content holds promise for scalable medical education, but evidence comparing AI and clinician-produced podcasts remains limited. We investigated whether AI-generated podcasts achieve comparable student satisfaction to clinician-recorded podcasts, enabling broader curriculum coverage without overburdening clinical educators, and explored students' attitudes towards AI-generated podcasts. METHODS: Medical students were randomised into two groups, with group 1 receiving clinician-produced podcasts on type 1 diabetes (T1DM) and AI-generated podcasts on type 2 diabetes (T2DM). Group 2 received AI-generated podcasts on T1DM, and clinician-produced podcasts on T2DM. Students were blinded to the podcast types. The outcome measures included students' satisfaction (measured using the Student Satisfaction with Educational Podcasts Questionnaire (SSEQ)) and students' attitudes to AI using an open-ended survey. A Bonferroni correction was applied (α = 0.025). RESULTS: = 165), 56% reported production method mattered, citing the sound of the voice, engagement of speaker, and ethical concerns as major issues with AI podcasts. DISCUSSION: AI-generated podcasts achieved non-inferior satisfaction to clinician-recorded podcasts. However, students' psychological preference for human teaching and negative perceptions of AI pose a barrier to scalable benefits, necessitating efforts to build trust and acceptability.
No takes yet. Share an insight, caveat, or question.
Guha et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: