Michael Rowe

Trying to get better at getting better

AI in health professions education: Moving beyond binary thinking

We’re facing a pivotal moment in HPE where AI is rapidly entering our classrooms, simulation labs, and clinical teaching environments. The discourse around AI integration in health professions education often falls into two simplistic camps: “AI is destroying learning” versus “AI will revolutionise everything.” Neither position serves us well.

The false dichotomy

Saying that AI is bad for learning is like saying humans are bad for learning because some teachers use ineffective methods. This framing misses a fundamental truth: good pedagogy is good pedagogy, regardless of who—or what—facilitates it.

When we dismiss AI categorically, we’re being intellectually lazy. Instead of examining specific implementations and their alignment with sound educational principles, we’re allowing fear or unfamiliarity to drive our decision-making.

The health professions context

In health professions education, this conversation takes on additional complexity:

  1. Clinical reasoning development: AI tools can provide students with varied case presentations and immediate feedback on diagnostic reasoning, but without proper scaffolding, they might short-circuit the struggle that builds expertise.
  2. Professional identity formation: Learning to be a healthcare professional involves more than knowledge acquisition—it requires role modelling, reflection, and community. How do we ensure AI complements rather than replaces these human elements?
  3. Patient safety: The stakes in health professions education are uniquely high. How do we leverage AI while ensuring students develop the vigilance and responsibility needed for patient care?

Principles for AI Integration in health professions education

Instead of looking for all the ways that AI fails in the education context, let’s instead focus on how it can be implemented effectively:

  • Alignment with learning objectives: AI tools should serve clearly defined educational goals, not be used simply because they’re available.
  • Appropriate scaffolding: As with any teaching method, AI-enhanced learning requires thoughtful scaffolding that gradually transfers responsibility to the learner.
  • Authentic assessment: If we’re concerned about AI enabling shortcuts, we should design assessments that evaluate applied knowledge and skills in authentic contexts.
  • Ethical framework: Students should understand both the capabilities and limitations of AI in healthcare, developing critical perspectives on when and how to use these tools.

Moving forward together

As a community of health professions educators, we have both the responsibility and opportunity to shape AI integration in health professions education. Rather than outsourcing this critical conversation to technologists or administrators, we must actively engage with these tools, evaluate their impacts, and share our findings.

The question isn’t whether AI belongs in health professions education—it’s already here. The real question is how we can harness its potential while preserving the human connection, clinical wisdom, and ethical foundation that define excellence in healthcare.


Share this


Discover more from Michael Rowe

Subscribe to get the latest posts to your email.