Are AI scribes useful tools in medical education, or a crutch that imperils learning?

August 3, 2026 - 09:53

Medical educators are increasingly wrestling with a new question in the classroom: should future doctors be trained to use AI scribes that listen to patient visits and write the clinical note? The answer is not simple. On one hand, these tools save time and reduce the clerical burden that drives many physicians to burnout. On the other, there is a real fear that handing the note-taking to a machine too early could weaken the very skills that make a good doctor.

The core of the debate is about what happens during the act of writing a note. For decades, the process of listening to a patient, sorting through their story, and typing up a summary was seen as a way to force a trainee to think clinically. It was a time to practice picking out the key details, forming a differential diagnosis, and documenting the reasoning. If a student simply accepts an AI-generated note, that mental workout disappears. The worry is that residents will become skilled at editing text but less skilled at synthesizing information on their own.

Proponents of using AI scribes in training argue that the skill of the future is not handwriting notes but managing information. They point out that physicians already spend too many hours on screens, and that off-loading the mechanical part of note writing frees up time for actual patient interaction. They also note that AI notes are not perfect; they still require careful review for accuracy, which is a skill in itself. The goal, they say, is to teach students how to use the tool well, not to ban it.

Still, many educators are pushing for a phased approach. They suggest that early students should write their own notes without AI, to build the foundation. Only later, once the basics are solid, should they be introduced to AI scribes as an aid, not a replacement. The key is to make sure that the technology does not become a crutch that prevents learners from ever developing the habit of deep clinical listening. The question is not whether to use AI, but when and how, so that it enhances training instead of quietly eroding it.