Exploring nurse educators' and students' use of large language models for academic purposes in a developing country:...

To explore how nurse educators and students use Large Language Models (LLMs) for academic purposes and their perceived challenges.

Empirical investigation of how nurse educators and students use LLMs is limited, particularly in developing countries.

A mixed method design with an embedded approach.

A convenient sample of nurse educators and students reported their socio-demographic characteristics and experiences with LLMs, and a purposive sub-sample was interviewed to elicit reasons for non-use or details and challenges of using LLMs. Descriptive statistics were used for quantitative data, and qualitative data were analyzed using inductive content analysis.

The participants (n = 165) included 56 faculty members, 35 graduate students and 74 undergraduates. Those who were aware of LLMs but had not used them (n = 34) reported four reasons: limited awareness and familiarity; preference for traditional resources; fear of perceived lack of authenticity; and lack of trust and privacy concerns. Participants who had used LLMs (n = 81) reported their uses and challenges faced. Reported uses included: writing assistance and education assistance, with some participants reporting unethical uses, such as answering online exams. Challenges included difficult use, inaccuracy of responses, overly advanced content and limited capabilities.

This study showed the potential of LLMs in nursing education and revealed unethical practices. Awareness campaigns, training programs, clear institutional policies and improving the reliability of LLMs are required. Although the study focused on nursing students and educators in a developing country, many of the insights are likely transferable across other disciplines and regions.