Modernizing a Classic: An Artificial Intelligence–Powered Patient Simulator in Athletic Training Education

Introduction/Background Simulation is central to athletic training education, but standardized patients and high-fidelity mannequins can be costly and logistically complex. We developed an artificial intelligence–based patient simulation using a large language model to increase access to authentic clinical encounters while maintaining educational rigor. Framework The Patient Case Simulator generative pretrained transformer is grounded in experiential learning, cognitive apprenticeship, and transformational learning. It primarily targets professional communication skills, such as patient interviewing, empathy, and consent seeking, while supporting clinical reasoning development appropriate for novice learners. These outcomes align with clinical decision-making models in athletic training education. Implementation We customized a generative pretrained transformer within the ChatGPT (OpenAI) platform using deidentified case materials, detailed patient personas, and safety parameters. Educators can begin with a library of 3–5 cases spanning common athletic training domains. After brief behavior testing to ensure role fidelity and realistic challenge, the simulator can be deployed in guided lab settings, asynchronous practice, or remediation. Text- and voice-based interactions allow students to practice communication and reasoning skills, followed by structured debrief prompts to facilitate reflection. Discussion/Implications This technique reduces logistical barriers to simulation, enables scalable and repeatable communication practice, and complements existing instructional strategies. Ongoing evaluation should examine learner outcomes, perceived realism, and communication performance using validated assessment tools. Broader adoption may enhance access to experiential learning across athletic training programs while maintaining patient safety and educational integrity.