Key Insights for the Ethical and Appropriate Use of Artificial Intelligence by Medical Learners.

INTRODUCTION

The rapid advancement and adoption of large language models (LLMs) in various academic domains necessitate an examination of their role in scholarly works by medical learners.This paper seeks to discern the implications of LLM use by medical learners when preparing works for publication. While LLMs possess great potential to revolutionize the academic writing process, they can detract from the learning process when used by students and residents who are still learning how to research, formulate ideas, and write cohesive arguments.

MATERIALS AND METHODS

An environmental scan of both traditional evidence-based sources and gray literature was performed to glean best practices of generative AI in medical education. Sources included peer-reviewed journals, open-source websites, and previous publications in this field ranging from 2015 to 2023.

RESULTS

We propose several strategies to detect AI involvement: direct inquiry to the learner, assessing the coherence level of the content in contrast to the learner's known capabilities, recognizing patterns of shallow insight or depth, utilizing plagiarism and AI-specific detection tools, and monitoring for fabricated citations-a known pitfall of LLMs.

CONCLUSIONS

Although LLMs offer potential efficiencies in academic writing, unchecked use can jeopardize the development of essential critical thinking and analytical skills in medical learners. Ultimately, mentors and primary investigators are responsible for ensuring learners are advancing and appropriately utilizing new and emerging technology. This study provides a foundational framework for educators in both responsible use of generative AI and best practices.