Medical Education Leaders Talk AI Integration, Faculty Adoption, and Key Trade-offs

Artificial intelligence (AI) is transforming how students learn, how faculty teach, and how institutions govern themselves. A May 19 panel discussion for the Technology Innovation in Medical Education Series (TIME Series) brought together three medical education technology leaders to discuss how their institutions are using AI to support learner development.

Moderated by Kathleen Ludewig, MSI, MPP, co-director of the Yale School of Medicine (YSM) educational technology and innovation team, panelists Paul Lawrence, Mount Sinai Health System & Icahn School of Medicine at Mount Sinai; Teggin Summers, PhD, Stanford Health Care and School of Medicine; and Michael Campion, MEd, University of Washington School of Medicine, explored the ways in which their institutions are integrating AI into the curriculum.

Ludewig started the discussion by asking each panelist to share their institution’s flagship AI projects or initiatives.

At the University of Washington, Campion described the recent launch of an AI tutor using Google NotebookLM. The tutor was initially designed by a faculty member as a study tool for a specific course at one of the school’s six teaching sites.

Recognizing the tool’s potential, the university expanded the tutor’s reference material to include a wide range of curated, faculty-selected content and opened access to students at all sites. The AI is structured to create multiple choice questions and engage students in tutoring to help with concept retention and studying.

Stanford developed an AI-powered virtual patient platform that allows students to interact with simulated patient cases, receive feedback on their diagnoses, and practice clinical interactions. The tool is used in Stanford’s Practice of Medicine curriculum and gives students a structured, low-pressure way to build clinical judgment.

At Mount Sinai, the initial focus was to establish AI governance and form the “AI Committee on Teaching, Learning, and Discovery.” Lawrence commented that having the foundational governance has been critical in coordinating AI pursuits across education, research, and clinical settings . This structure has helped them quickly adopt products like OpenAI’s ChatGPT Edu platform and ensure their faculty have the tools and resources needed to keep up with rapid AI technology changes.

A common theme among panelists was the importance of getting their communities comfortable, curious, and excited about using AI.

For Stanford, this has meant adding AI-focused roundtables and discussion blocks to annual retreats and recruiting AI champions to host workshops and create training materials for the wider community.

Stanford also developed an “AI Foundations” course which is mandatory for everyone in the information technology group and has recently been opened to the larger Stanford Medicine community. Additionally, they developed a secure GPT platform, which allows their community to experiment and explore AI in a safe environment.

Lawrence spoke about the valuable partnership with their university library in educating the Mount Sinai community. He commented that libraries are a natural educational hub, not only for faculty but for students and researchers, and their library’s foundational AI training courses have helped the community get comfortable with AI tools.

Lawrence also emphasized the importance of governance and how their AI framework has helped faculty know where to focus their efforts when exploring the technology.

Campion talked about the University of Washington’s monthly interest groups, where people can learn how their colleagues are using AI and explore possible areas for collaboration. The institution also held an in-person faculty and staff retreat where colleagues showcased different AI projects and innovations happening around campus.

Although the panelists mostly focused on AI successes and positive developments, AI adoption has not been without challenges.

Both Lawrence and Summers commented on AI pricing models and how expensive tools may lead to developmental inequities among smaller medical schools and universities. Lawrence noted that some institutions may be better equipped to negotiate contracts financially, administratively, or as part of a consortium, which can deepen the technological divide already being seen across sectors.

For Summers, it’s not just the financial cost, but also the time and effort spent negotiating, maintaining, and developing these tools.

She said there is a great opportunity to collaborate not only across departments but also across institutions and systems, and leveraging what’s already been developed before investing time and resources in recreating the wheel.

There’s also a feeling of “franticness” because of how quickly AI technology is developing. Summers commented that there's a pervasive urgency to catch up or be proactive, but it’s important they consider how the student experience is and will be affected by AI.

Campion noted that they know AI will change what it means to be a doctor, but what they don’t know is how.

Perhaps the most powerful takeaways from this discussion were the commonalities among each institution’s triumphs and challenges. As Summers mentioned, the time, effort, and cost required to understand and develop these tools is significant.

This panel emphasized the importance of cross-institutional conversations and highlighted opportunities to create a more sustainable, equitable, and environmentally conscious AI landscape.

Email Innovate.MedEd@yale.edu to request access to the webinar recording.