Impact of Artificial Intelligence on academic competence in medical research: A Scoping Review

The effect of Artificial Intelligence on scientific research is currently controversial (over-use, refusal). The objective of this scoping review was to summarize the impact of Generative Pre-Trained Bots (ChatGPT) on medical research, for optimal use.

In accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines, three electronic databases (MEDLINE, Web of Science, and Scopus) were searched for publications on the use of ChatGPT in medical research up to September 30, 2024. The inclusion criteria encompassed systematic reviews, meta-analyses, and reviews published in English or French. A synthesized and consensual perspective was then derived using the SWOT (Strengths, Weaknesses, Opportunities, and Threats) approach.

Among the 120 articles identified during the study period, 33 publications were reviewed to describe the impact of ChatGPT on medical research skills. It emerged that ChatGPT was considered a tool for generating innovative ideas and analyzing "Big Data," helping researchers execute manually time-consuming processes. However, its responses may be accompanied by "hallucinations," posing a threat to the scientific integrity of medical research.

ChatGPT serves as a valuable aid in medical research, particularly in its conceptualization and writing phases. However, its potential drawbacks, such as "hallucinations," highlight the need to strengthen young researchers' skills in the proper use of Artificial Intelligence.