AI Clinic Builds Empathy Skills
At the University of Arizona, a new kind of outpatient clinic is opening doors, virtually. The AI Outpatient Clinic, developed through a collaboration between the Center for Transformative Interprofessional Healthcare (CTIPH), UITS iDX, and Campus Web Services (CWC), is helping health sciences students strengthen one of the hardest-to-teach skills in medicine: empathy.
The platform hosts 21 virtual patient cases, each designed to simulate complex, real-world scenarios where students must listen actively, respond compassionately, and navigate sensitive topics such as stigma and substance use. Using artificial intelligence and natural language processing, the clinic provides learners with consistent, low-risk opportunities to practice patient interviewing, no faculty facilitator required.
“Traditional simulations can be limited by time, access, or consistency,” said project collaborators Margie Arnett (CTIPH), Joe Arnett (CWC), and Jay Timsina (UITS iDX). “The AI Outpatient Clinic bridges that gap by offering scalable, repeatable practice that’s available anytime, anywhere.”
Built with healthcare-specific data and frameworks such as the Interprofessional Education Collaborative (IPEC) competencies and the Quintuple Aim for health equity, the system gives students automated, formative feedback on their performance. That includes evaluation of empathy, rapport-building, and interprofessional awareness – metrics aligned with four validated communication and collaboration tools.
During early testing, students submitted more than 900 queries, with results showing strong engagement and increased confidence in handling difficult conversations.
Faculty are beginning to adopt the Virtual Outpatient Clinic in courses, underscoring
its potential to address gaps in traditional training by providing consistent, feedback-rich opportunities to practice empathy-driven care.
The next phase, launching in spring 2026, will involve a multi-institutional research study exploring health professions students’ perceptions, satisfaction, and performance in using the AI Outpatient Clinic to simulate patient encounters and develop key communication skills.
The project team emphasized that building an educational AI tool requires more than technical skill. It requires intentional design. “Uncontrolled AI can cause institutional damage if it provides inaccurate or inappropriate responses,” the team noted. “By training our AI assistant with healthcare-specific context, we ensure realism, reliability, and respect.”
The AI Outpatient Clinic now serves as part of a broader ArizonaAI ecosystem, connecting academic innovation across disciplines. For the university, it reflects a growing commitment to integrating AI responsibly into learning – where technology not only teaches knowledge but also nurtures compassion.
Content for this article was adapted from the poster “An AI Outpatient Clinic: Scalable Training for Empathy and Interviewing,” presented by Margie Arnett, Joe Arnett, and Jay Timsina at the 2025 IT Summit.