Skip to main content
Canadian Medical Education Journal logoLink to Canadian Medical Education Journal
. 2025 Dec 22;16(6):52–53. doi: 10.36834/cmej.82035

Enhancing equity in medical school interview prep with ChatGPT

Améliorer l’équité dans la préparation aux entrevues d’admission en médecine avec ChatGPT

Kyra Yuan Li 1,, William Tien Tran 2, Farah Hashemi-Sabet 3, Fok-Han Leung 4
PMCID: PMC12826817  PMID: 41584934

Introduction

An interview is one of the last and most critical steps that every applicant faces towards medical school admission. Many applicants turn to preparatory courses, mentors, and AI programs designed to support them. However, those expensive resources are difficult to access for students from a low socioeconomic status (SES). Research has shown that when students from low SES backgrounds are provided access to such resources, they report improved confidence and feel they expanded their knowledge base.1 As such, it is important to consider other avenues for better accessibility and equity. The Voice feature from OpenAI’s ChatGPT allows for free verbal interview practice and is able to return immediate personalized feedback for each interaction.2 A study found that medical students using ChatGPT as practice for an interview scored significantly higher than those who did not, suggesting this tool’s potential effectiveness as practice for medical school interviews.3 Based on my experience with using ChatGPT and its Voice feature, this commentary investigates the feasibility of the tool and explores different ways of using it.

Practicing with ChatGPT voice feature

I started using ChatGPT’s Voice feature to provide examples of generic interview questions. As a result, I felt increasingly comfortable with the AI and strengthened my responses and communication skills.

Please give me a practice interview question and then give feedback on my response.

Eventually, I asked for specific questions that I wanted to practice. This approach helped me target areas where I felt less confident, allowing me to improve the quality of my answers:

Pretend you are an interviewer for a medical school asking me about my strengths and weaknesses, then give me feedback on my response.

With situational interview questions, ChatGPT was able to generate different scenarios, and provide feedback on my responses. This was useful in learning different perspectives and exploring ways to approach a problem.

Please give me an example of a conflict scenario and then ask me how I would resolve it and give feedback on my response.

I used the Voice feature to answer the prompts and chat-based interactions when reviewing feedback and asking follow-ups. ChatGPT provided feedback on the overall content, perspective, tone, and structure of my response. To make the feedback more relevant to me, I uploaded a PDF containing a list of my personal experiences. This allowed ChatGPT to suggest better integration of those experiences into my answers and recommend other relevant experiences I could utilize. The AI provided feedback in a respectful, unbiased, and encouraging tone, going through both my strengths and areas for improvement.

It was also convenient when analyzing previous practice sessions, as I could copy the transcript ChatGPT made of our interaction into a separate document. While the initial immediate feedback was helpful, I often asked follow-up questions to get more in-depth recommendations.

Advantages

OpenAI can provide practice as an accessible, cost-effective alternative to costly commercial resources. If individuals have access to a device and reliable internet connection, they would be able to use it. Additionally, it is available at any time, allowing for flexibility in scheduling, which can be useful for students who need to balance various responsibilities, such as jobs and classes. Furthermore, by giving the AI a PDF containing previous experiences or personal details, it can generate more personalized feedback and practice.

Limitations

Interviews usually require the applicant to show their face, and ChatGPT does not have a free video feature available to practice with. As such, the AI cannot provide feedback on how interviewees present themselves visually, such as facial expression or hand movements. Additionally, humans are the ones ultimately reviewing the interview, and AI may not be able to fully replicate the experience of human interaction. It is also important to consider and think critically about the responses being provided, as the information and feedback may not always be applicable or credible in every scenario.

Conclusion

For me, ChatGPT was a viable alternative as a resource for medical school interview practice. It can serve as practice, and give immediate, personalized feedback to users. While it is more accessible, there are still limitations to keep in mind, such as the lack of video. If possible, it would be ideal to use this tool in combination with other forms of practice and resources to make up for any gaps. Further research to test its feasibility and performance edge both alone and in concert with other resources is important.

Conflicts of Interest

There are no conflicts of interest and no funding.

Authorship

All authors have each signed their own attestation statement that they meet the requirements of authors. The lead author, in addition to their own attestation as an author, has signed an attestation that all authors listed on this paper meet the requirements for authors.

Edited by

Marcel D’Eon (editor-in-chief)

References


Articles from Canadian Medical Education Journal are provided here courtesy of University of Saskatchewan

RESOURCES