Over 60% of GP trainees already use ChatGPT for reflective practice — according to Learner+'s own pilot data. The question is not whether AI will be used for reflections (it already is) but whether a purpose-built tool is better than a general-purpose LLM.
ChatGPT for Reflections
ChatGPT is powerful, flexible, and free. You can describe a clinical encounter and ask for a structured reflection. The output is fluent and well-organised. But it carries specific risks for GP trainees.
Information governance: ChatGPT is a consumer tool with no clinical data assurances. Entering patient details — even anonymised ones — into a US-hosted general-purpose LLM raises IG concerns that your deanery may not accept. Learner+ is designed for clinical context with appropriate data handling.
Generic output: ChatGPT produces reflections that sound polished but often lack clinical specificity. "I reflected on the importance of patient-centred care and plan to improve my communication skills" is generic. "I identified that I did not explore the patient's ideas about statin side effects, which led to a missed opportunity for shared decision-making per NICE CG181" is specific. Purpose-built prompts produce specific output.
No FourteenFish integration: ChatGPT output must be copied and pasted into FourteenFish manually. Learner+ integrates directly.
No curriculum alignment: ChatGPT does not know the RCGP curriculum capabilities or GMC revalidation standards. Learner+ is designed around them.
Learner+ for Reflections
Learner+ (by CMEfy) provides AI-generated reflection prompts tailored to clinical encounters, with direct FourteenFish integration and alignment to RCGP competencies. The pilot data (24 GP trainees, 4-6 months): 77% found reflections easier, 90.5% would recommend, average usability 4.27/5.
The limitation is availability — Learner+ is early-stage, and access beyond the pilot cohort may be limited.
The Principle
AI should scaffold reflection, not generate it. The learning comes from your own thinking — the AI should make the process of capturing and structuring that thinking easier, not replace it. ARCP panels may scrutinise AI-generated reflections, and overreliance undermines the developmental purpose of reflective practice.
Where iatroX Fits
iatroX's CPD module offers AI-assisted reflection scaffolding mapped to professional domains — purpose-built for clinical learning documentation with appropriate clinical data handling. The content of your reflection must be genuinely yours; the tool should just make the process easier.