Generative AI is raising new questions for clinical educators about how to evaluate tools thoughtfully, how to support learners navigating these technologies, and how to think about AI's role in clinical training. There are no settled answers yet, and most programs are working through these questions in real time.
This page is intended as a starting point. It brings together newsletters, research tools, policy frameworks, and guidance on academic citation practices that may be useful as you begin exploring AI integration in your curriculum. It's a working list, and we'll update it as the field develops.
What are the leading AI newsletters and journals in medicine?
Here are a few sources worth bookmarking to stay up to date:
- AI Breakfast — Daily digest; good for staying current across the field
- Ground Truths — Dr. Eric Topol — Deep, evidence-grounded analysis from a leading clinical AI researcher
- One Useful Thing — Dr. Ethan Mollick — Practical applications; consistently relevant for educators
- NEJM AI's "This Week" — Peer-reviewed clinical AI research with editorial commentary
- JAMA+ AI — JAMA's AI channel; covers both research and policy
What do professional organizations say about AI in clinical education?
Major healthcare organizations have begun publishing formal guidance, though this area continues to develop. The documents below represent some of the most relevant frameworks for clinical educators:
- NAM AI Code of Conduct — Key principles for responsible, human-centered AI across health and medicine (2025)
- AMA Principles for Augmented Intelligence — Governance framework for AI development, deployment, and use in clinical care
- AMA AI Literacy Policy — New 2025 policy expanding AI training across medical school and CME
- ACP Position on AI in Clinical Care — Internal medicine–specific guidance on AI in diagnosis and treatment
- WHO Ethics & Governance of AI for Health — Global ethical framework; 2025 update addresses generative AI and LLMs
- AAMC Principles for Responsible AI in Medical Education — Seven principles for ethical AI use in medical education (Version 2.0, July 2025)
- AAMC Advancing AI Resource Collection — Free, practical repository of AI tools and frameworks for medical schools and faculty
- AAMC AI Competencies for Medical Educators — Defines what faculty need to know and do with AI; mapped to 2025 international framework
What are the best AI research assistant tools for clinical educators?
For literature review, evidence synthesis, and citation management:
- Elicit — AI-assisted systematic review and research synthesis
- Scite.ai — Citation analysis; flags supporting vs. contradicting evidence
- Semantic Scholar — Free AI-powered research search from Allen Institute
- Connected Papers — Visual citation mapping; useful for literature reviews
How to cite AI in academic writing: APA guidance on citing ChatGPT and generative AI — The current APA standard to share with faculty and learners
Who should I follow to stay current on AI in healthcare?
A short list of trusted voices to follow on X/Twitter and LinkedIn:
- Ethan Mollick — AI in education; practical and evidence-grounded
- Ethan Goh, MD — Stanford physician-scientist; rigorous clinical trials on LLMs in real patient care
- Eric Topol, MD — Cardiologist and Scripps Research director; leading clinical AI researcher
- Vivian Lee, MD, PhD, MBA — Former Verily/Google Health executive; focuses on AI, health system transformation, and equity
FAQ
What's the difference between general-purpose AI tools and a clinical readiness platform like DDx? General-purpose tools like ChatGPT can generate cases and provide feedback, but require significant faculty effort to validate, structure, and score consistently. Purpose-built platforms like DDx provide a faculty-developed case library, structured rubric assessment, and longitudinal analytics — with far less faculty overhead per learner interaction.
Can I use public AI tools like ChatGPT for clinical education without a privacy concern? For educational purposes using de-identified or fictional cases, public tools are generally appropriate. Any use involving real patient data requires institutional review and should route through your health system's compliant platforms. Check your institution's current AI use policy before deploying tools in a patient care context.
How do I evaluate whether an AI-generated clinical case is accurate? AI-generated cases should always be reviewed by a subject matter expert before use in a formal curriculum. Hallucination — the generation of plausible but incorrect clinical information — remains a real risk. Treat AI output as a first draft, not a finished product.
