Transforming Case Write-Up Evaluation with AI: Insights from a Pilot Study at a Private Medical College

Abstract In this study, authors have explored use of AI tools in enhancing evaluation of undergraduate case writeups to determine uniqueness of the student-generated research questions. Case write-ups are used as continuous assessment method in department of Family Medicine at Shifa College of Medicine, Islamabad. The task for students involves documenting patients’ medical information and developing research questions based on their respective cases. Work of 50 students from two consecutive batches rotating in Family Medicine was assessed. The originality of the submitted questions was assessed using GPT Zero, to identify AI generated material. To ensure academic integrity, plagiarism checks were also carried out using Google Scholar and Semantic Scholar to find any overlaps with already published material. From a sample of 150 research questions, nearly half of all the research questions (42.7 %) were plagiarized, 34% of questions were AI generated, and 23.33% were original questions.