In the era of generative AI tools like ChatGPT, which handle 230 million weekly health queries, medical education faces a critical gap. The AI-PACE Framework is a groundbreaking model from researchers Scott P. McGrath, Katherine K. Kim, Karnjit Johl, Haibo Wang, and Nick Anderson. It provides a structured roadmap for embedding generative AI literacy across all training stages.
This framework ensures physicians master AI from basics to leadership, revolutionizing AI in medical education.
What is the AI-PACE Framework?
The AI-PACE Framework provides a structured approach to integrating artificial intelligence, including generative AI, into medical education across the full training continuum.
AI-PACE expands Bloom’s Taxonomy into four pillars: Psychomotor, Affective, Cognitive, and Embedded (PACE), tailored for generative AI applications in healthcare.
Key Features Explained:
- Cognitive Domain: Builds foundational knowledge in generative AI algorithms, large language models (LLMs), and probabilistic reasoning—essential for interpreting AI outputs in diagnostics and treatment planning.
- Psychomotor Domain: Hands-on skills like validating generative AI-generated reports during clinical workflows, using “human-in-the-loop” verification.
- Affective Domain: Cultivates trust calibration, ethical decision-making, and bias awareness through generative AI failure case studies.
- Embedded Domain: Delivers spiral learning from undergraduate medical education (UME) to continuing medical education (CME), avoiding one-off workshops.
This generative AI-optimized medical education framework aligns with Harden’s integration ladder, making AI a core competency for every physician, not just specialists.
AI-PACE Implementation: Stage-by-Stage Generative AI Training Guide
The framework maps generative AI skills to training milestones, using practical tools such as OSCEs, simulations, and AI ethics projects.
Training Stage:
- UME (Preclinical)- AI Basics & Ethics. LLM prompt engineering in evidence-based medicine modules; bias detection exercises.
- GME (Residency)- Clinical Integration. Real-time generative AI validation in rotations; “AI-Fail” case debriefs.
- CME (Ongoing)- Leadership & Oversight. Selecting/auditing generative AI tools; developing institutional policy.
Pro Tip: Start with generative AI pilots in med school curricula to boost student AI readiness by 40%, per early adopters.
Why AI-PACE Matters for Generative AI in Healthcare Education
As technology changes medical education, it will reshape healthcare from drafting patient notes to predicting outcomes; traditional curricula lag behind. AI-PACE addresses this by prioritizing:
- Generalist Skills: Equips all doctors to evaluate generative AI, beyond radiology or pathology.
- Ethical Guardrails: Tackles hallucinations and inequities in LLMs.
- Longitudinal Learning: Ensures lifelong AI proficiency amid rapid advancements.
Pioneering med schools are already adopting similar models, signaling a shift to AI-augmented medical training.
Future of Generative AI Medical Education with AI-PACE
Researchers urge the development of faculty programs and rigorous pilots to scale AI-PACE globally. By fostering patient-centered, HIPAA-compliant, and ethical use of AI, this framework positions medical education for the generative AI revolution in healthcare.
Ready to implement? Download the full AI-PACE paper: arxiv.org/abs/2602.10527
Share your thoughts on generative AI in med ed below!
