OpenAI announced ChatGPT for Clinicians today — a free tool for individual doctors, nurse practitioners, physician assistants, and pharmacists. Led by Nate Gross, OpenAI's Head of Health, this is the third OpenAI health product in four months, joining ChatGPT Health (7 January, consumer-facing) and ChatGPT for Healthcare (8 January, enterprise). Three products, three audiences, maximum market confusion. This article cuts through it.
The Three-Product Breakdown
ChatGPT Health (7 January 2026) — The Consumer Product
ChatGPT Health is patient-facing. Users connect Apple Health data, medical records (US EHR via FHIR integration), and wellness apps including MyFitnessPal, Peloton, Function, AllTrails, and Weight Watchers. They can then ask wellness questions, request plain-language explanations of lab results, compare insurance options, and prepare for appointments with their physician.
Conversations within ChatGPT Health are encrypted and isolated — explicitly not used to train foundation models. The product was developed with input from over 260 physicians across 60 countries over two years, incorporating more than 600,000 feedback interactions on model outputs. OpenAI reports approximately 40 million people use ChatGPT daily for health-related questions — ChatGPT Health channels that existing behaviour into a purpose-built experience with additional safety rails.
The safety evidence is mixed. A Nature Medicine study published on 23 February 2026, conducted by researchers at the Icahn School of Medicine at Mount Sinai (Ramaswamy et al.), found that ChatGPT Health under-triaged 52% of gold-standard emergencies. Patients with diabetic ketoacidosis and impending respiratory failure were directed to "24-48 hour evaluation" rather than the emergency department. Classical emergencies (stroke, anaphylaxis) were correctly triaged 100% of the time — but ambiguous or atypical emergencies were missed in more than half of cases.
UK availability: explicitly excluded. OpenAI's launch documentation states eligibility for "users with ChatGPT Free, Go, Plus, and Pro plans outside of the European Economic Area, Switzerland, and the United Kingdom." OpenAI cited regulatory considerations — UK GDPR classifies health data as special category data with strict processing requirements that OpenAI chose not to address for the initial launch.
ChatGPT for Healthcare (8 January 2026) — The Enterprise Product
Built for hospital-wide deployment. Powered by GPT-5 models optimised for clinical workflows and evaluated through HealthBench (physician-written rubrics prioritising safety behaviours — knowing when to refuse, when to escalate) and GDPval. HIPAA-compliant with customer-managed encryption keys and business associate agreements. Supports institutional policy integration so clinical outputs align with hospital-specific protocols.
Already deploying at AdventHealth, HCA Healthcare, Boston Children's Hospital, Cedars-Sinai Medical Center, Memorial Sloan Kettering, Stanford Medicine Children's Health, and UCSF. Includes citations from peer-reviewed studies with title, journal, authors, and publication date — moving beyond ChatGPT's usual conversational output toward verifiable clinical evidence synthesis. Enterprise pricing with custom quotes — individual clinicians cannot purchase it.
UK availability: enterprise only, no UK NHS Trust deployments announced. The product is designed around US healthcare infrastructure — HIPAA compliance, US EHR integration, US institutional procurement processes.
ChatGPT for Clinicians (22 April 2026 — Today) — The Individual Product
Launched today. Free for individual verified clinicians — doctors, nurse practitioners, physician assistants, and pharmacists. This product fills the gap between the consumer tool (patients self-triaging, with documented safety limitations) and the enterprise tool (which requires institutional IT deployment and enterprise pricing).
Details are still emerging — the announcement is hours old. Verification likely requires NPI (National Provider Identifier), which is a US-specific identifier that UK clinicians do not have. Based on the consistent pattern of US-first rollout across all three products, UK availability at launch is unlikely — though OpenAI has not explicitly confirmed geographic restrictions for this product yet.
The key question: does ChatGPT for Clinicians include the citation features from ChatGPT for Healthcare, or is it essentially standard ChatGPT with clinician verification? The distinction matters enormously for clinical utility and safety.
The UK Position: Three Products, Zero Available
As of today, UK clinicians cannot access any of OpenAI's three health products in their intended clinical form.
ChatGPT Health: geo-blocked from the UK. ChatGPT for Healthcare: enterprise-only with no UK institutional customers. ChatGPT for Clinicians: appears US-first with NPI-based verification.
The exclusion pattern is not arbitrary. The UK regulatory environment creates specific barriers that OpenAI has not yet addressed. The MHRA's National Commission into the Regulation of AI in Healthcare has been actively developing recommendations — the call for evidence closed in February 2026. Any clinical AI tool seeking formal deployment in UK healthcare would likely need to address MHRA medical device classification (if the tool provides clinical decision support, it may meet the definition of a medical device under UK regulations), UK GDPR compliance for health data processing, and clinical safety standards including DCB 0129.
These are not trivial requirements. They represent months to years of regulatory work — work that UK-native platforms have already completed.
What UK Clinicians Can Use Right Now
The UK is not a clinical AI desert — it just cannot access OpenAI's products yet. Purpose-built alternatives exist, and some have already navigated the regulatory pathway that OpenAI has not attempted.
iatroX is available now, free, and is a UKCA-marked, MHRA-registered Class I medical device. Ask iatroX retrieves and synthesises NICE guidelines, CKS summaries, peer-reviewed literature, and SmPC data — the reference standards UK clinicians actually use. The platform also includes 15+ adaptive exam Q-banks covering PLAB, UKMLA, MRCP, MRCGP AKT, MRCEM, PSA, MSRA (free), and specialist diplomas including DRCOG, DFSRH, DGM, DipIMC, FFICM, DTM&H, and GPhC CRA (£99/year via /boards). Plus 80+ clinical calculators with editorial content, evidence summaries, and guideline references. This combination — clinical AI, exam preparation, calculators, and CPD within a single regulated device — is unique in the UK market.
OpenEvidence is free for verified clinicians globally. It searches peer-reviewed medical literature using proprietary models (not ChatGPT), is HIPAA-compliant and SOC 2 Type II certified. Approximately 15 million clinical consultations per month. Recently embedded into Mount Sinai's Epic EHR. US-evidence-centric but globally accessible — a strong research synthesis tool with no UK guideline specificity.
Medwise AI offers UK-focused enterprise clinical AI with NHS Trust deployments, integrating local Trust policies and formularies alongside national guidelines. Enterprise licensing required — not available to individual clinicians.
The Regulatory Gap Is Substantive
iatroX has navigated the UK regulatory pathway: UKCA-marked, MHRA-registered, DCB 0129 clinical safety governance with systematic hazard identification and risk controls, post-market surveillance obligations. ChatGPT's health products have no medical device registration in any jurisdiction.
This gap is not bureaucratic theatre. Medical device registration requires the manufacturer to systematically identify clinical hazards, implement risk controls, maintain a clinical safety case, and conduct ongoing post-market surveillance. These are the mechanisms that catch safety issues before they reach patients — the mechanisms that flagged the risks a Nature Medicine study later confirmed.
When ChatGPT for Clinicians eventually reaches the UK, it will need to address the same requirements that iatroX has already met. The regulatory pathway takes years. iatroX has already completed it.
What This Means for the Market
OpenAI entering clinical AI with a free individual-clinician product compresses the entire market. Every clinical AI tool now competes against "free from OpenAI" — regardless of whether that free tool is actually available to UK clinicians or appropriate for UK clinical practice.
Purpose-built tools must differentiate on dimensions that a general-purpose model cannot easily replicate. UK guideline specificity — NICE NG28 is not ADA guidelines, and the dose, threshold, and management pathway differences between US and UK practice are clinically significant. Regulatory status — MHRA registration and UKCA marking provide clinical governance assurance that unregistered tools cannot. Workflow integration — exam preparation, clinical calculators, CPD documentation are capabilities that ChatGPT does not and will not offer.
iatroX's position: MHRA registration + UK guideline depth + exam preparation + clinical calculators in one free platform. Available now. Purpose-built for UK practice. This combination is the moat.
Try Ask iatroX — free, available now, UK guideline-aligned, MHRA-registered →