Large Language Models in Ophthalmology Scientific Writing: Ethical Considerations Blurred Lines or Not at All?

Abstract

To discuss the implications of large language models (LLMs) in ophthalmology research, as well as the associated ethical considerations. Perspective. This discussion reviews the potential uses of LLMs such as ChatGPT in ophthalmology research, highlights the associated threats and ethical considerations, and proposes solutions for the use of LLMs in ophthalmology research and scientific writing. With the increasing interest in LLMs, such as ChatGPT, their diverse utility has been widely explored, including their application in research and scientific writing. LLMs have the potential to guide researchers throughout the different stages of their research, from idea generation to drafting a scientific piece. However, there are significant ethical concerns and challenges related to scientific integrity in ophthalmology research that should be addressed by scientific journals. Our review of the 10 highest-impact-factor ophthalmology journals revealed that the number of journals addressing this topic in their submission guidelines is rapidly increasing. Therefore, we propose certain domains that all journals should consider regarding the use of LLMs in research. As LLMs continue to improve, their use in scientific writing will remain a contentious issue due to the ethical dilemmas involved in determining the appropriate scope of their use. This article reviews the ethical dilemmas related to the use of LLMs in ophthalmology research and calls for the prompt development of guidelines for their ethical use in manuscript writing as ophthalmology journals update their editorial policies with respect to LLMs.