Answers are grounded in this paper and cite their sources.
Introduction Training future physicians occurs in a context where preventable clinical errors remain frequent and safe decision-making is increasingly complex. Clinical simulation has therefore become a key strategy to promote learning without risk to patients. More recently, artificial intelligence (AI) has begun to shape simulated environments by enabling more interactive, adaptive, and personalised experiences. However, empirical evidence regarding the educational implications of AI-based clinical simulation for undergraduate medical students remains fragmented. Methods We conducted a scoping review following Joanna Briggs Institute guidance and the PRISMA-ScR checklist. Empirical studies published between 2015 and 2025 (English or Spanish) were included if they evaluated AI-based clinical simulation among undergraduate medical students and reported learning outcomes and/or cognitive processes. Searches were performed in PubMed, Scopus, Web of Science, and BIREME (LILACS and SciELO), complemented by backward and forward citation tracking. Titles/abstracts and full texts were screened using predefined eligibility criteria. Database searching identified 777 records, from which 21 studies met the eligibility criteria. Backward and forward citation tracking identified nine additional eligible studies, resulting in a final sample of 30 included studies. Results The included studies primarily examined AI-enabled virtual patients (including NLP-based systems and language models), conversational simulations (chatbots/LLMs), and platforms providing automated feedback; fewer addressed immersive environments (e.g., VR) or social robotics. Frequently reported outcomes included history-taking, communication skills, OSCE-type performance, clinical reasoning, and decision-making. Many studies described favourable findings, particularly when interventions incorporated deliberate practice and structured or real-time feedback. However, study designs and outcome measures were heterogeneous. Conclusions The current body of literature suggests that AI-based clinical simulation may support undergraduate medical training, particularly in areas related to clinical reasoning and communication, by facilitating repeated practice and scalable feedback opportunities. Nonetheless, the evidence remains methodologically diverse and largely limited to short-term evaluations. Careful curricular integration and critical appraisal of implementation are therefore warranted.
No takes yet. Share an insight, caveat, or question.
Rincón et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: