Generative AI as the First Patient: Practice, Feedback, and Confidence

Generative artificial intelligence (GenAI) offers new potential for medical education by providing personalized learning opportunities at scale that address current faculty shortages, such as limited access to clinical preceptors, yet its integration into formal medical training and acceptability by students remains largely unexplored. We deployed a large-language model-based AI Patient Actor app in a first-year medical curriculum, enabling students to conduct simulated clinical interviews while receiving immediate feedback. Students rated their experiences as overwhelmingly positive, with comfort using GenAI in general increasing significantly after repeated engagement. Sentiment analysis of student comments showed appreciation for the AI approaches' flexibility, low-pressure environment, and real-time feedback that students felt helped with clinical reasoning and communication skills. Students identified limitations regarding the authenticity of AI interactions compared to human encounters. This study shows that AI patient actors are perceived as acceptable educational adjuncts in early clinical training, particularly in formative environments where they provide students with immediate feedback and greater autonomy over their learning process, though they cannot fully replace interactions with human actors and patients.

The online version contains supplementary material available at 10.1007/s40670-025-02473-x.