Generative artificial intelligence in medical education: from knowledge assessment to clinical reasoning and professional competence

Artificial intelligence in medical education: a narrative review across four functional domains.

BACKGROUND Artificial intelligence (AI) is increasingly reshaping medical and health-professions education through adaptive tutoring, generative content creation, simulation analytics, automated assessment, and diagnostic-reasoning support. Since 2023, large language models and multimodal AI systems have expanded AI from relatively narrow analytic tools into interactive educational agents capable of dialogue, feedback, and content generation. OBJECTIVE This narrative review asks: How does AI contribute to medical education across four core educational functions, and what methodological, ethical, and implementation constraints should guide responsible integration? To answer this question, we synthesize evidence across four functional domains: AI-supported instruction and content generation, AI-augmented simulation and procedural training, AI-supported assessment and learning analytics, and AI-assisted diagnostic reasoning and clinical cognition. METHODS A structured search of PubMed, Scopus, and Web of Science (January 2018-January 2025), supplemented by citation tracking, identified 540 records, of which 90 met the eligibility criteria. Evidence was analyzed using an iterative qualitative synthesis approach that combined inductive coding of educational mechanisms with deductive organization into the four predefined functional domains. Findings were summarized descriptively because of substantial heterogeneity in interventions, outcomes, and study designs. RESULTS In instruction and content generation, AI-supported tools were associated with preliminary gains in knowledge acquisition, learner engagement, and self-regulated study, although accuracy depended on supervision and prompt quality. In simulation and procedural training, computer-vision, motion-analytics, and VR/AR systems supported immediate feedback and were associated with improved procedural efficiency in early studies. In assessment and learning analytics, AI tools reduced feedback latency and improved scoring consistency, but validity, fairness, and explainability remained central concerns. In diagnostic reasoning, AI-supported case platforms and LLM-based dialogue improved short-term reasoning performance and metacognitive calibration, while evidence for long-term clinical transfer remained limited. CONCLUSION AI appears most defensible as an augmentative educational partner that strengthens feedback, personalization, and competency-based progression, rather than as an autonomous substitute for educators. The current evidence supports cautious, supervised implementation, but remains early-phase and methodologically uneven. Multicenter validation, transparent reporting, equity-focused implementation, and structured AI-literacy training are essential before AI can be integrated more broadly into high-stakes educational workflows.

Artificial Intelligence in Pharmaceutical and Health Sciences Education: 

Artificial intelligence (AI) increasingly influenced pharmaceutical and health sciences education by transforming teaching, assessment, simulation-based learning, and curriculum development. This PRISMA-guided narrative literature review evaluated current educational applications, examined associated ethical challenges, and identified future directions for responsible curricular integration. A structured search of PubMed, Scopus, Web of Science, Google Scholar, and ScienceDirect identified English-language publications from 2020 to 2026. After duplicate removal, title and abstract screening, and full-text eligibility assessment, 32 studies and policy documents met the predefined criteria and were included in a qualitative narrative synthesis. Data were extracted and synthesized thematically. The review showed that AI-driven systems supported adaptive learning, virtual patient simulations, learning analytics, automated assessment, generative AI, and large language models, improving personalized learning, student engagement, educational efficiency, and clinical training. Important challenges involved data privacy, algorithmic bias, academic integrity, transparency, equitable access, AI literacy, and institutional governance. International policy recommendations emphasized ethical implementation and responsible governance in higher education. Overall, AI demonstrated considerable potential to improve pharmaceutical and health sciences education when combined with evidence-based educational practices, comprehensive faculty and student training, and robust ethical and regulatory frameworks. Continued interdisciplinary collaboration and responsible governance remained essential to ensure that AI complemented, rather than replaced, human-centered teaching and learning.

Artificial Intelligence in Medical Education: Implications for Undergraduate, Postgraduate, and Continuing Training in Family Medicine

Artificial intelligence (AI) is being increasingly utilized in medical education, with growing interest in its potential to enhance learning, assessment, and faculty support across undergraduate, postgraduate, and continuing medical education. While much of the literature focuses on specialty training or technology-driven innovation in isolation, family medicine represents a distinct educational domain characterized by its breadth, clinical uncertainty, multimorbidity, longitudinal care, prevention, and shared decision-making. These features present specific educational challenges and opportunities for integrating AI-supported tools. This narrative review synthesizes peer-reviewed literature on the application of AI in medical education, specifically within the context of family medicine. This review utilized PubMed and a consensus search engine to identify relevant reviews, empirical studies, and articles that address AI-supported learning, simulation training, competency-based education, workplace assessment, faculty support, and educational governance. The evidence was integrated through a narrative synthesis, guided by the Scale for Assessment of Narrative Review Articles. The literature indicates that AI can facilitate personalized learning pathways, augment simulation-based training, and aid in the integration of longitudinal assessment data, particularly in educational contexts characterized by distributed supervision and diverse learner requirements. There is a paucity of evidence concerning the application of AI in high-stakes assessments, autonomous decision-making, and its long-term impact on professional identity formation. In all areas, the successful incorporation of AI necessitates robust educational governance, faculty oversight and alignment with fundamental educational principles. AI should be considered an augmentative educational technology that can enhance but not supplant human-centered teaching, supervision, and professional judgment. Future research should prioritize the validation of educational outcomes, the clarification of ethical safeguards, and the establishment of best practices for the responsible integration of AI into family medicine education.

Generative AI as a catalyst for next-generation medical education: innovation, ethics, and implementation

From concepts to clinicals: A comparative study of artificial intelligence platforms for competency-based medical education and postgraduate pharmacology problem-solving

Artificial intelligence (AI) platforms have increasingly integrated into medical education, offering innovative avenues for enhancing critical thinking, subjective learning, and clinical reasoning. Large language models such as ChatGPT, Gemini, and Meta AI simulate human-like responses and are increasingly being utilized as adjunctive educational tools. However, systematic evaluations comparing their performance within competency-based medical education (CBME) in pharmacology remain limited. A cross-sectional comparative study was conducted to evaluate ChatGPT, Gemini, and Meta AI in answering 55 higher-order pharmacology questions mapped to CBME undergraduate competencies and postgraduate standards. Responses were anonymized and independently scored by three blinded pharmacology faculty members for accuracy, completeness, and clarity. Statistical analysis included the Friedman test, Wilcoxon signed-rank test for pairwise comparisons, and repeated measures analysis of variance (ANOVA). Data visualization was performed using bar graphs and box-and-whisker plots. ChatGPT achieved the highest mean score (8.5 ± 1.0), followed by Gemini (7.9 ± 1.2) and Meta AI (6.4 ± 1.5). The Friedman test indicated a highly significant difference among the platforms ( χ ² = 96.69, P < 0.001). Post hoc Wilcoxon signed-rank tests showed that ChatGPT significantly outperformed both Gemini and Meta AI ( P < 0.001), while Gemini also outperformed Meta AI ( P < 0.001). Repeated measures ANOVA confirmed significant differences in mean scores across platforms ( F = 217.36, P < 0.001). Among the evaluated AI platforms, ChatGPT demonstrated superior performance in solving higher-order pharmacology questions aligned with CBME frameworks. These findings support the integration of AI tools, as adjunctive resources in pharmacology education.

Generative artificial intelligence and the epistemology of medical learning: from cognitive augmentation to epistemic dependency

Generative artificial intelligence (GAI) is rapidly transforming medical education by expanding access to information, enabling personalized learning, and improving measurable educational outcomes. Systematic reviews, meta-analyses, and experimental studies consistently report gains in learner satisfaction, communication abilities, practical skills, and short-term academic performance following GAI-assisted educational interventions. Yet these benefits have largely been evaluated through performance-centered metrics, while the cognitive and epistemological implications of algorithmically mediated learning remain insufficiently examined. This Perspective explores how GAI may be reconfiguring the conditions under which medical knowledge is acquired, interpreted, and integrated. Drawing on medical epistemology, philosophy of science, educational theory, and evidence-based medicine, we argue that the educational significance of GAI extends beyond efficiency and information accessibility. By increasingly participating in explanatory, interpretive, and inferential activities, generative systems are redistributing forms of cognitive work that have traditionally contributed to the development of autonomous clinical judgment. We propose that the relationship between GAI and medical learning is best understood as a continuum that extends from cognitive augmentation to epistemic dependency. Along this continuum, algorithmic systems may expand opportunities for learning and reasoning while simultaneously altering the intellectual dispositions through which critical evaluation, uncertainty management, and independent judgment are cultivated. The principal challenge for contemporary medical education is therefore not whether future physicians will learn alongside intelligent systems, but how educational environments can harness the benefits of AI-mediated learning while sustaining the cognitive capacities that underpin responsible clinical reasoning. Cultivating epistemological resilience may become a central objective of medical education in the age of GAI.