OpenAI has built a consumer-to-professional-to-enterprise healthcare stack in four months. No company has attempted this pace of healthcare AI product deployment before. Understanding the strategy — and its gaps — matters for every clinician evaluating which tools to adopt.
The Three-Tier Stack
Tier 1 — Consumer: ChatGPT Health (7 January 2026). Patients manage their own health. Connect medical records (US EHR via FHIR), Apple Health, and wellness apps (MyFitnessPal, Peloton, Function, AllTrails, Instacart, Weight Watchers). Ask wellness questions, get plain-language lab explanations, prepare for appointments. OpenAI reports approximately 40 million daily US health queries. Developed with 260+ physicians across 60 countries. Over 600,000 feedback interactions over two years. Health chats are encrypted, isolated, and not used for foundation model training. UK/EEA/Switzerland explicitly excluded. Medical record integrations US-only.
Tier 2 — Individual Clinician: ChatGPT for Clinicians (22 April 2026 — today). Free for verified individual doctors, nurse practitioners, physician assistants, and pharmacists. Led by Nate Gross, OpenAI's Head of Health. This product fills the critical gap between the consumer tool (where patients self-triage, with documented safety limitations) and the enterprise tool (which requires institutional IT deployment and enterprise pricing that excludes individual clinicians). Details are still emerging. Verification likely requires NPI — suggesting US-first availability.
Tier 3 — Enterprise: ChatGPT for Healthcare (8 January 2026). Hospital-wide deployment. HIPAA-compliant with customer-managed encryption keys, business associate agreements, and institutional policy integration. Powered by GPT-5 models (GPT-5.2 specifically cited). Evaluated via HealthBench (physician-written rubrics prioritising safety behaviours) and GDPval. Already deploying at AdventHealth, HCA Healthcare, Boston Children's Hospital, Cedars-Sinai, Memorial Sloan Kettering, Stanford Medicine Children's Health, and UCSF. Includes citations from peer-reviewed studies. Enterprise pricing — custom quotes, institutional IT deployment.
The Technical Foundation and Its Limits
GPT-5 models optimised for healthcare. OpenAI's internal benchmarks claim GPT-5.2 achieves 100% diagnostic accuracy on some standardised scenarios and 98.7% on healthcare-specific tests. Built with 260+ physicians across 60 countries and 600,000+ feedback interactions over two years.
But internal benchmarks and independent evaluation tell different stories. The Nature Medicine study (Ramaswamy et al., Mount Sinai, February 2026) independently evaluated ChatGPT Health and found: 52% undertriage of gold-standard emergencies, 35% overtriage of non-urgent presentations, significant anchoring bias when social context minimised symptoms (OR 11.7, 95% CI 3.7-36.6), and inconsistent suicide-crisis safeguard activation.
The gap between internal benchmarks and independent evaluation is a consistent pattern across AI — not unique to OpenAI. But it is precisely why independent evaluation and regulatory frameworks matter. A manufacturer claiming their product is safe is not the same as an independent regulator verifying that it is. This distinction is the foundation of medical device regulation.
The Competitor Landscape
The clinical AI market that OpenAI is entering is not empty. Each competitor occupies a distinct position.
OpenEvidence. The largest clinician-facing clinical AI platform globally. $12 billion valuation following a $250 million Series D in January 2026. Approximately 15 million clinical consultations per month. Free for verified clinicians worldwide — ad-funded with pharmaceutical advertising during loading screens (CPMs reportedly $70-150+). Embedded into Mount Sinai's Epic EHR in March 2026 — a landmark integration that places AI clinical decision support inside the physician's primary clinical workflow. Uses proprietary models (not ChatGPT). HIPAA-compliant, SOC 2 Type II certified. US-evidence-centric but globally accessible. Compare OpenEvidence vs iatroX →
iatroX. UK-focused. MHRA-registered, UKCA-marked Class I medical device. Free. NICE/CKS/literature/SmPC-grounded clinical AI via Ask iatroX. 15+ adaptive exam Q-banks across all major UK exams. 80+ clinical calculators. CPD documentation tools. The only platform combining clinical AI + exam preparation + calculators + CPD in a single regulated device — purpose-built for UK clinical practice. Try iatroX →
Medwise AI. UK enterprise clinical AI. NHS Trust deployments with local policy and formulary integration alongside national guidelines. Enterprise licensing — not available to individual clinicians. Strongest for institutional workflow integration. Compare Medwise vs iatroX →
Glass Health. Differential diagnosis generation from clinical presentations. US-focused. VC-backed. Narrower scope but focused execution on its specific use case.
Doximity DoxGPT. Available to US physicians through the Doximity network (2 million+ US physicians). PeerCheck citation verification. Integrated with Doximity's physician communication and professional networking platform — reaching clinicians where they already spend professional time online.
Microsoft Copilot for Healthcare. Enterprise only. Six-figure annual commitments. Requires Microsoft ecosystem integration. Dragon Copilot heritage for clinical documentation alongside broader administrative workflow automation.
The UK Gap — Challenge and Opportunity
None of OpenAI's three products are available to individual UK clinicians in their intended clinical form. The UK regulatory environment — MHRA medical device classification requirements, UK GDPR health data protections, MHRA National Commission developing AI healthcare recommendations — creates barriers that OpenAI has not yet addressed.
This creates a paradox: the UK market is underserved by the world's largest AI company, but well-served by purpose-built UK-native tools. iatroX has already navigated the regulatory pathway — UKCA marking, MHRA registration, DCB 0129 clinical safety governance — that any entrant, including OpenAI, would need to complete. When ChatGPT for Clinicians eventually reaches the UK, iatroX will already have regulatory status, UK guideline depth, and clinical trust that takes years to build from scratch.
What This Means for the Market
Free clinician-level AI from the world's largest AI company compresses prices and raises expectations across the entire market. Purpose-built tools must differentiate on dimensions a general-purpose model cannot easily replicate.
Guideline specificity. UK vs US vs global. NICE NG28 is not ADA guidelines. The dose, threshold, and management pathway differences between jurisdictions are clinically significant. Global models need local guideline layers — iatroX has this for the UK already.
Regulatory status. MHRA/UKCA registration versus no registration. This is a trust signal for clinicians, for employers, and increasingly for regulators. As the MHRA's National Commission formalises AI healthcare recommendations, registered platforms have a head start.
Workflow integration beyond chat. Exam preparation, clinical calculators, CPD documentation — capabilities that ChatGPT does not and will not offer. These create daily touchpoints that a chat-only clinical AI cannot replicate. A clinician who prepares for MRCP on iatroX, uses its calculators during ward rounds, and consults Ask iatroX during consultations has integrated the platform into their professional workflow in a way that no general-purpose AI can match.
The winners will be tools that clinicians trust enough to use during the 10-minute GP consultation — not the ones with the most parameters or the largest marketing budget.