Press Release

Study reveals medical students’ use of ChatGPT in education and calls for ethical guidelines

Researchers from the Keck School of Medicine of USC found that a majority of U.S. medical students are actively using ChatGPT for academic purposes, highlighting both its potential benefits and ethical concerns in medical education.

Sheila Rodriguez November 13, 2024
Closeup of young woman using her computer to take notes

Photo/iStock

A recent study conducted by researchers at the Keck School of Medicine of USC sheds light on the widespread use of generative AI models, particularly ChatGPT, among medical students in North American medical colleges. The study, conducted in May 2023, surveyed 415 students from 28 medical schools in May 2023 to gauge their views on and their current use of ChatGPT and similar technologies. 

The results revealed that 96% of respondents had heard of ChatGPT, with 52% reporting its use for medical school coursework.

The most common uses reported by students included seeking explanations of medical concepts, assisting with diagnosis and treatment plans, and proofreading academic research.

Students found ChatGPT particularly beneficial for studying, writing, and clinical rotations, showing a hopeful attitude toward its future integration with existing study resources. These findings, published in PLOS Digital Health, underscore the tool’s perceived value among students, particularly in its ability to save time and enhance productivity.

Calls for ethical guidelines and regulatory frameworks

The study also uncovered concerns about the limitations of ChatGPT. Students reported encountering vague or inaccurate responses and bias in AI algorithms, with 75% of users experiencing issues using the tool.  

Additionally, ethical considerations such as potential patient privacy violations (e.g., sharing sensitive patient information with the AI) and risks of plagiarism (e.g., using AI-generated content without proper attribution) were noted, emphasizing the need for strict regulatory guidelines endorsed by 92% of respondents.

“We recognize the potential of AI in transforming medical education, but clear rules are essential to prevent misuse,” said Conner Ganjavi, a fourth-year medical student at the Keck School of Medicine of USC and lead author of the study. 

“Understanding students’ perspectives is crucial in developing workable instructional courses, guidelines, and regulations that ensure the safe, productive use of generative AI in medical education,” said Giovanni Cacciamani MD, director of the AI Center at USC Urology.  

Insights into the Medical Students’ Perspectives

The researchers crafted an open online survey to gather insights from a diverse group of medical students.

Most respondents were female (61.7%) and between 25–34 years old (55.2%). The highest proportion of respondents were in their first year (35.4%), followed by third year (28.4%), second year (6.5%), and fourth year (9.2%). Regarding specialty interest, 55.4% were interested in internal medicine or medicine sub-specialties, while 35.9% were interested in surgery or surgical sub-specialties, 23.1% were interested in primary care, and 39.8% were interested in other specialties like pediatrics and emergency medicine. 

The methodology was adapted from previous surveys and underwent a rigorous two-round iterative process to ensure the clarity and relevance of the questions. Researchers thoroughly tested the final version of the survey to ensure proper functionality and logical question structuring to best capture the students’ perspectives.

About this research

In addition to Ganjavi, the study’s other authors are Michael Eppler, Devon O’Brien, Lorenzo Storino Rmacciotti, and Giovanni Cacciamani from the Keck School of Medicine of USC; Muhammad Shabbeer Ghauri from the California University of Science and Medicine, Colton, California; Issac Anderson from the Wayne State University School of Medicine, Detroit, Michigan; Jae Choi from the UT Southwestern Medical School, Dallas, Texas; Darby Dwyer from the Texas A&M School of Medicine, Bryan, Texas; Claudia Stephens from the Frederick P. Whiddon College of Medicine, University of South Alabama, Mobile, Alabama; Victoria Shi from the University of Missouri-Kansas City School of Medicine, Kansas City, Missouri; Madeline Ebert from the Medical College of Wisconsin, Milwaukee, Wisconsin; Michaela Derby from the Sanford School of Medicine, University of South Dakota, Vermillion, South Dakota; and Bayan Yazdi from the Stritch School of Medicine, Loyola University Chicago, Maywood, Illinois.