AI as a Teaching Assistant: Exploring Chatbots in Medical Curriculum Support

Artificial intelligence (AI), particularly in the form of conversational chatbot systems, presents promising opportunities for transforming medical education by offering personalized, scalable, and interactive learning experiences. This systematic review evaluated the effectiveness, feasibility, and acceptability of AI chatbots functioning as teaching assistants in medical curricula. A total of 16 peer-reviewed studies published between 2015 and 2025 were analyzed, covering diverse educational settings across undergraduate and postgraduate medical training. Studies were sourced from PubMed, Scopus, Web of Science, Google Scholar, and the Cochrane Library, applying PRISMA guidelines. Inclusion criteria focused on English-language research involving medical students, residents, or educators using AI chatbots for curriculum support. Quantitative outcomes showed significant improvement in academic performance (+12%), knowledge retention (+18%), and learner satisfaction (mean rating 4.3/5). Acceptance rates among users ranged from 75% to 92%, reinforcing the broad receptivity toward chatbot-assisted learning. Qualitative findings emphasized personalization, accessibility, and educator efficiency, as chatbots reduced repetitive instructional tasks and enabled more time for mentorship. The study also identified challenges, including institutional readiness, educator resistance, and concerns about data privacy and loss of human interaction. Review duration was approximately ten months, encompassing search development, data screening, extraction, and synthesis. These results affirm the pedagogical value of AI chatbot systems in improving engagement and supporting learner autonomy while underscoring the need for robust infrastructure and ethical safeguards. This review contributes actionable insights for the future integration of AI chatbots into clinical education, advancing technologically enriched teaching methods without compromising core humanistic values.