A new sector-wide index and a fast-approaching regulatory deadline, are putting a number on whether UAE clinics can be found and understood by the AI assistants patients are already using. Dr Andrey Perfilyev, CEO and co-founder, Yma Health, looks at the issue and provides some solutions.
A growing share of patients no longer begin their search for care with a search engine or a phone call. They begin it with a question typed into a conversational AI assistant: which clinic treats this condition, which doctor is available, whether a procedure is covered by insurance, and what it might cost. The assistant then answers using whatever public information it can find about the provider. If that information is thin, outdated or missing, the assistant cannot give a confident answer, and the patient may never reach the clinic that could have treated them.
This shift is no longer marginal, and nowhere is it clearer than in the UAE itself. The 2026 Edelman Trust Barometer’s special report on trust and health found that 59% of UAE residents already use AI to manage their health, against a global average of 35%, one of the highest adoption rates of any market surveyed. OpenAI has reported that health is among the most common subjects people raise with conversational AI worldwide, accounting for more than five per cent of all ChatGPT messages and touching roughly one in four of its weekly active users, with most of those conversations happening outside normal clinic hours. Separate research from Gallup found that recent AI-health users in the United States often consult AI both before and after a doctor’s visit, and some use the advice as a substitute for a visit altogether. What sets health apart from most other things people ask a chatbot is that it sits inside one of the most tightly regulated, clinically sensitive and practically consequential categories an AI system can be asked to answer, raising the stakes of getting the underlying public information right.
Deadline for business to adopt agentic AI
A regulatory clock is already running alongside that behavioural shift. In May 2026, Dubai’s Crown Prince Sheikh Hamdan bin Mohammed gave the emirate’s entire private sector, healthcare included, a two-year deadline to adopt agentic AI, backed by Chamber of Commerce training tracks, incubators and dedicated funding, following an April 2026 federal directive to deliver half of all UAE government services through autonomous AI agents by 2028. Abu Dhabi has moved in parallel: the Department of Health’s Unified Medical Operations Command Centre already uses AI agents for real-time incident detection and emergency-response coordination, described by its chairman as giving the system the equivalent of forty times its human workforce. For providers, AI-mediated discovery and AI-mediated operations are converging on the same two-year horizon, making public-data readiness a near-term operational question rather than a distant one.
To answer that question with evidence rather than anecdote, Yma undertook a systematic assessment of public provider data across the UAE, published as the UAE Healthcare AI Readiness Index 2026 and authored by Daniel Gusev, Head of Strategy at Yma Health. The exercise deliberately does not assess clinical AI adoption, electronic health record maturity or diagnostic capability. It measures something more specific and actionable: whether the public information a clinic publishes is complete and structured enough for an AI system to answer a patient’s practical questions about it.
The research began with a working dataset of 6,732 UAE healthcare-adjacent records, gathered and checked between May and June 2026. Of these, 4,082 qualified as core providers: clinics, hospitals, medical centres and polyclinics with an identifiable public presence. Each record was assessed against 14 observable data fields, including a website, phone and email contact, WhatsApp availability, a description of services, a doctor roster, listed specialties, indicative pricing, insurance information, languages spoken and public review evidence, each weighted by its importance to a patient’s decision. A provider’s overall score is the share of the available 24 weighted points its public data covers.
A visible sector, but not always an answerable one
The headline result is a national average of 79.8 out of 100, suggesting a healthcare sector with reasonably strong public visibility. The detail tells a more uneven story. A website or official page was identifiable for 83.2% of providers, and a structured list of services or specialties for close to nine in ten. Beyond that baseline, the picture weakens. A named doctor roster, the kind of information a patient needs to judge whether a specific clinician is available, was published by only 61.4% of providers. A working WhatsApp channel, increasingly the preferred route for patients to take the next step after an AI-generated recommendation, was found for 41.6%. Indicative pricing, one of the most requested pieces of information in patient enquiries, was published by just 7.1%.
Read together, these figures describe a sector that is visible but not always answerable. A provider can appear confidently in an AI-generated response and still leave the assistant unable to say which doctor performs a procedure, whether a treatment is covered by insurance, or roughly what it costs. For a patient using an AI assistant precisely to avoid the friction of calling around, an incomplete answer is often functionally equivalent to no answer at all.
Wide variation between emirates
The Index also found meaningful geographic variation. Dubai, with 2,173 core providers in the sample, recorded the strongest combined performance across scale, website coverage, doctor answerability and pricing visibility, at an average of 83.8 out of 100. Abu Dhabi, with 746 providers, trailed by roughly seven points, driven largely by a lower rate of published doctor rosters and WhatsApp access. Sharjah, Al Ain, Ajman, Ras Al Khaimah, Fujairah and Umm Al Quwain showed further variation, generally correlating with the density and maturity of the local provider market. The pattern suggests that AI-search readiness, like earlier waves of digital adoption, will spread unevenly across the country unless it is deliberately measured and managed at a sector level.
What the score does, and does not, measure
It is worth being precise about what a score of 79.8 does and does not mean. It reflects the share of weighted public evidence available in Yma’s model, not how often a clinic is actually recommended inside any individual AI assistant, nor a judgement on clinical quality or data accuracy. Some gaps are also legitimate: a provider may withhold pricing or route WhatsApp enquiries through a governed contact centre for regulatory reasons, and the Index does not penalise sound operational judgement. What it does capture is the raw material any AI system needs to answer a patient’s question, a useful and largely unmeasured input into how healthcare access functions in an AI-mediated market.
From readiness to referrals: what clinics should do now
Search engine optimisation remains important, but patients increasingly see a synthesised answer before any list of links. Generative engine optimisation and answer engine optimisation extend, rather than replace, good SEO: they help AI systems discover, interpret, verify and cite provider information. Current guidance points to several practical priorities for clinics and hospital groups:
- Create an authoritative source of truth. Maintain crawlable, individual pages for every doctor, location and service, with current specialties, credentials, languages, insurance networks, availability and contact details, not buried in PDFs or images.
- Structure and substantiate the information. Use structured data that matches the visible content, and keep healthcare information patient-centred, medically reviewed and attributed to qualified professionals.
- Publish decision-making details. Where regulation and policy allow, include indicative prices, eligibility criteria, preparation requirements, expected recovery and answers to common patient questions.
- Keep every source consistent. Align the website, directories, Google Business Profiles, insurer listings and review platforms; conflicting details weaken both patient trust and AI confidence.
- Move from answers to actions. Every recommendation should lead to a monitored booking form, WhatsApp channel or phone line. Clinics might also explore Model Context Protocol (MCP), an emerging connector that lets authorised AI agents check live availability and, with proper consent, complete a booking directly.
- Measure and govern visibility. Assign a named owner, set a review cadence, and test realistic patient queries across major AI assistants, tracking citations, referral traffic, booking completion and generative-AI visibility reporting, not rankings alone.
For regulators, the opportunity sits at sector level: a repeatable methodology can track public-data readiness by emirate, specialty and service type over time, distinguishing genuine gaps from information withheld for sound reasons, and informing minimum public-information standards without requiring any change to clinical practice itself.
Patients are already asking AI assistants the questions that used to be spread across search engines, referrals, front desks and follow-up calls, and Dubai’s adoption clock means AI-mediated discovery and AI-mediated intake will both be standard practice within two years, not optional extras. Whether assistants can answer well, and whether that answer converts into a booked appointment, depends entirely on the public information providers choose to publish and how quickly they prepare to respond once an AI-generated answer sends a patient their way. The UAE Healthcare AI Readiness Index 2026 is offered as a baseline for that work, to be repeated as the market and the underlying AI systems evolve.
Access the report
The report – “The UAE Healthcare AI Readiness Index 2026” – can be accessed here:
https://www.yma.health/uae-ai-readiness-index
About Yma
Yma is a UAE healthcare AI platform focused on AI-assisted healthcare discovery and patient access. Yma Layer helps healthcare providers and groups understand how they are represented in the public data used by AI systems, identify answerability gaps and prioritise improvements. Yma Workspace supports the response, routing, follow-up and booking workflows that follow discovery. More information: www.yma.health
About the author Dr Andrey Perfilyev is CEO and co-founder of Yma Health (Reforma Health FZ-LLC), a Dubai-based healthcare AI platform. He is a medical doctor and health-technology entrepreneur with more than 15 years of experience in digital health, preventive medicine and personalised care, and has led two prior company exits
