Updated 10 November 2026.
Clinicians are using AI for medical questions. The GMC's own research confirms it — over a quarter of UK doctors have used some form of AI in practice. The question is no longer whether AI is being used but which tool is safest and most useful.
This article compares the four AI tools clinicians most commonly reach for — ChatGPT, Claude, Perplexity, and iatroX — across the dimensions that matter for clinical use.
The Comparison Framework
Accuracy. Does the tool give the correct clinical answer?
Citations and provenance. Does it show where the answer came from? Can you verify?
UK guideline grounding. Does it reference NICE, CKS, SIGN, and current UK medicines information — the sources that govern UK practice?
Hallucination risk. How often does it generate plausible but incorrect information?
Cost and access. Is it free? Does it require verification or institutional access?
Regulatory status. Is it classified as a medical device or clinical tool?
Additional clinical features. Does it offer learning, reasoning support, or CPD?
ChatGPT (OpenAI)
Name the product and its enabled tools. General chat, cited research and a professional healthcare workspace are materially different configurations. OpenAI’s current clinician product includes clinical search, research and documentation support. This article does not establish a measured accuracy or hallucination rate for that current experience.
For a UK task, inspect source relevance, access and the actual workflow rather than transfer historical benchmark percentages or model names to the present product.
Claude (Anthropic)
Claude can be used through general chat or configured healthcare tools. Anthropic’s healthcare offering includes organisational workflows and a development route; its documented ecosystem includes clinical information access. Assess the enabled sources, intended use and local implementation.
This comparison does not establish that Claude is more accurate, better calibrated or safer than another current product. Those claims require a specified evaluation.
Perplexity
What it is: An AI-powered search engine that combines LLM generation with real-time web search and citation.
Accuracy: Variable. Perplexity searches the web and synthesises answers with inline citations. The accuracy depends on the quality of the sources it finds — which may include authoritative guidelines or patient-facing health content of variable quality.
Citations: Perplexity supplies linked web sources. Compare their statement-level support with the actual retrieved sources in the other configured products; the presence of links alone does not establish superior citation quality.
UK grounding: Variable. Perplexity retrieves sources through web search, so UK material may appear without being prioritised consistently. For UK questions, require citations to applicable NICE or CKS guidance and verify product-specific medicines information in the relevant SmPC on eMC.
Hallucination risk: Lower for factual claims (because it retrieves from web sources) but the sources themselves may be unreliable.
Cost: Free tier. Pro at $20/month.
Regulatory status: Not a medical device.
Additional features: None clinical.
Verdict: Assess the sources actually retrieved and their statement-level support. Linked web evidence is useful, but this article does not establish that Perplexity has better citation quality than every current clinical product.
iatroX
What it is: A UK-focused clinical AI using RAG over NICE, CKS and SIGN guidance, medicines information from relevant SmPCs on eMC, and peer-reviewed research.
Approach: UK clinical questions can trigger retrieval from configured sources before the model generates a synthesis. This can improve relevance and traceability, but the answer is still model-generated and must be checked against the cited material.
Citations: Answers can include links to retrieved NICE, CKS and SIGN guidance or medicines information from relevant SmPCs on eMC. Open the cited passage and verify its date, scope and applicability.
UK grounding: UK clinical context is a central iatroX design focus. Verify the relevant cited source and its scope; a product focus does not guarantee that every generated answer fits every patient or local pathway.
Hallucination risk: Source retrieval can improve traceability, but retrieval and synthesis errors remain possible. This article reports no controlled current comparison from which to assign iatroX a lower hallucination rate.
Cost: Ask and selected banks are free; wider learning access depends on the current plan. Check iatroX access.
Regulatory status: iatroX describes its UK reference-tool component as UKCA-marked and MHRA-registered. Educational tools are separately scoped. Registration is not a head-to-head accuracy result. Feature scope.
Additional features: Knowledge Centre for structured guideline browsing, Brainstorm for clinical reasoning, Q-Bank with spaced repetition for learning, CPD module for professional development.
Verdict: Start with iatroX for UK-source reference connected to question practice, Tutor and professional learning. Assess the complete learning route on a topic that matters to you.
Choosing a tool
For UK-source reference connected to question practice, Tutor and professional learning, start with iatroX. For broader research or record-related work, compare the exact professional workspace and enabled sources. Every consequential claim still needs appropriate verification.
Use the current evaluation framework to compare completed work rather than assign a universal winner.
Conclusion
A useful 2026 comparison names the product, source path and task. iatroX makes a clear case through UK-focused reference and connected learning; that case does not depend on treating competitors’ current clinical products as source-free chatbots.