JMIR Publications

ABSTRACT

Background:

Generative artificial intelligence (AI), particularly large language models (LLMs), has rapidly emerged as a potential enabler of personalized learning in medical education. In otolaryngology-head and neck surgery (ENT, Ear, Nose and Throat) education, a specialty characterized by broad knowledge demands and limited curricular exposure, generative AI has the potential to support personalized learning through adaptive assessment, tailored feedback, and scalable digital learning platforms. However, the scope, characteristics, and educational focus of existing studies have not been comprehensively mapped. Current evidence remains fragmented across benchmarking studies, surveys, and early educational interventions.

Objective:

This scoping review aimed to systematically map the literature on generative AI in otolaryngology education, summarize educational applications and performance characteristics, identify benefits and risks, and synthesize emerging directions for AI-enabled personalized learning platforms.

Methods:

We conducted a scoping review in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) criteria. PubMed and Web of Science were searched for English-language studies published between 2017 and 2025. Eighteen studies, including one preprint, met the inclusion criteria. Data were extracted into structured tables and analyzed using inductive thematic synthesis to identify major conceptual domains.

Results:

The included eighteen studies, originating from North America, Europe, and Asia, comprised benchmarking evaluations, survey-based adoption studies, educational interventions, and system development reports. Applications of generative AI ranged from knowledge retrieval, revision support, and board-style question practice to adaptive assessment systems and AI avatar–based induction programs. Seven major thematic domains were identified: (1) Educational Support and Learning Enhancement, (2) Learning Outcomes, Educational Effectiveness, and Pedagogical Value, (3) Model Performance, Accuracy, and Variability, (4) Assessment, Examination Readiness, and Question-Type Sensitivity, (5) Reliability, Safety, and Risks, (6) Adoption, Usability, and Implementation Barriers, and (7) Advanced AI Architectures and Future Educational Design. While newer-generation models demonstrated strong performance in structured tasks and some AI-supported systems improved learning outcomes, substantial limitations were consistently reported, including task-dependent accuracy, hallucinations, unreliable citation practices, and insufficient readiness for high-stakes assessment or clinical use.

Conclusions:

Generative AI shows considerable promise as a component of personalized learning platforms in otolaryngology education, but current systems remain unsuitable for unsupervised, high-stakes use. The most promising direction lies in hybrid, pedagogically governed architectures that integrate knowledge grounding, retrieval augmentation, and self-reflective verification. Future research should prioritize longitudinal educational outcomes, curriculum integration, and governance frameworks to enable safe and effective implementation. Clinical Trial: The search strategy registration was not required as our study did not involve human subjects.