JMIR Publications

ABSTRACT

Background:

The popularity of Large Language Models (LLMs) has grown exponentially across healthcare. Despite the wealth of literature on proposed applications within medical education, little research has investigated their real-world use.

Objective:

We aimed to determine their real-world benefits, facilitators, and barriers associated with the use of LLMs among medical students in the United Kingdom through semi-structured interviews.

Methods:

Semi-structured interviews guided by the Technology Acceptance Model, were conducted with 15 UK medical students from pre-clinical and clinical stages. An inductive thematic analysis was applied to interview transcripts to identify themes around actual system use, perceived usefulness, ease of use, and attitudes towards LLMs.

Results:

All participants used LLMs primarily for concise summaries and topic explanations, often favouring these tools over conventional search engines due to convenience and conversational ease. Many students were unaware of other potential applications proposed within the literature, such as simulation of clinical interactions. Significant concerns included inaccuracies and hallucinations, which decreased trust in outputs, especially among less experienced students. Additionally, students reported concerns about developing an overreliance on LLMs, potentially reducing deeper engagement with learning materials.

Conclusions:

LLMs have been widely adopted by UK medical students. There is therefore an urgent need for medical curricula to explicitly address not only effective use of AI technologies but also their broader educational implications, including assessment and curricula redesign. Future research should broaden geographical representation, investigate applications in low-resource settings, and integrate educators’ perspectives to establish future curricular guidance in an AI era. Clinical Trial: Not applicable