GPnotebook Pro subscribers now earn CPD for searching. Every AI Answer a Pro user reads is logged automatically to their CPD dashboard at 0.05 credits, equivalent to 3 minutes of learning, with no extra steps. It is a genuinely clever piece of product design, and it raises a question worth taking seriously: can every clinical search count as learning?
Why the idea is attractive
The insight behind the feature is sound. Clinicians learn constantly by resolving real patient-care questions during the working day, and almost none of that informal learning is ever recorded. Appraisal preparation then becomes an exercise in reconstructing what you learned months ago. Automatic capture attacks that problem directly: the learning event and its documentation happen in the same moment, and the administrative burden falls to near zero. This is CPD moving into the clinical workflow, which is exactly where it should live. The direction of travel is industry-wide: Wolters Kluwer announced in March 2026 that clinicians using UpToDate Expert AI can earn CME credit within the clinical workflow, so GPnotebook is in step with the category's most established players.
The educational distinction the model compresses
The difficulty is that a single automatic credit flattens a spectrum. Opening an answer, reading it, checking its source, reflecting on its relevance to the patient in front of you, and actually changing practice are five different activities with very different educational value. A fixed 3-minute credit treats them identically. It also treats questions identically: a simple factual check, a complex guideline review, and a repeat of a query you ran last week all log the same credit. GMC-facing CPD has always been anchored in reflection and impact rather than exposure time, and automatic capture records exposure.
To be fair to GPnotebook, the platform does not pretend otherwise. Reflective notes can be added, and the credit values are modest. But the incentive gradient of any automatic system is towards volume rather than depth, and clinicians should be conscious of it.
The wider GPnotebook Pro CPD machine
AI Answers slots into an established CPD infrastructure. Pro tracks page views, video views, podcast listens and quiz submissions; supports reflective notes; exports PDF reports for appraisal; and can sync reports to FourteenFish. As a passive-capture system it is arguably the most complete in UK primary care, and for clinicians who read GPnotebook daily it converts existing behaviour into appraisal evidence at essentially no cost.
How iatroX approaches the same problem
iatroX starts from the reflective end. The CPD workflow is built around the clinical question you actually asked, the learning point you took from it, and a short reflection, producing appraisal-ready output in which the clinician's reasoning is the record rather than a page-view log. Both philosophies are legitimate: GPnotebook emphasises frictionless automatic capture of reading activity; iatroX emphasises clinician-controlled reflection on clinical questions. An honest appraisal portfolio arguably wants both: broad evidence that you engage with information daily, and deep evidence that specific questions changed your practice.
Using automatic credits well at appraisal
For the clinician, the practical question is how to present automatically captured activity so it strengthens rather than pads a portfolio. A raw log of two hundred answer views is weak appraisal material on its own; the same log paired with a handful of written reflections is strong. A sensible pattern is to let the automatic capture run all year, then periodically pick the handful of questions that genuinely changed or confirmed your practice and attach a short reflection to each: what you asked, what you found, what you did differently. That converts exposure data into exactly the evidence appraisers are trained to look for, learning with demonstrable impact, while the automation quietly documents the breadth of your engagement underneath it. The credit value is the least important part; the searchable record of what you looked up, and when, is the genuinely useful asset.
What a stronger model would look like
The synthesis of the two approaches is not hard to sketch. Record the query automatically, so nothing is lost. Capture the sources actually reviewed, not just the answer viewed. Ask one lightweight question: did this change, or confirm, your practice? Let the clinician amend the credited duration when three minutes misrepresents the work. And export a concise reflection rather than a raw activity log. Each step keeps the friction low while restoring the element appraisers actually value, which is evidence of thought.
The fair conclusion
GPnotebook deserves credit here. Automatic CPD for AI Answers moves professional development into the flow of real clinical work, and the objections are refinements rather than rebuttals. The principle to hold onto is simple: automatic systems can record that learning probably happened; only the clinician can record what it meant. Use the automation, and keep the reflection yours.