Asia-Pacific patients are comfortable with AI in care. What should Australian practices publish?
Patients ask AI before they book. The provider facts AI should get right, and how to publish them within Australia's health advertising rules.

The short answer
Publish the facts patients check before booking, as plain text: services, practitioners and their registration, fees and rebates, telehealth, hours, referral rules and languages. Keep every claim inside Ahpra's advertising rules, with no testimonials and no outcome promises. AI answers then have accurate, permitted material to quote, and errors are easier to trace and correct.
Key takeaways
- In Ahrefs' study of 146 million desktop results from September 2025, AI Overviews appeared on 44.1% of medical queries, against 20.5% of results overall.
- Because medical questions so often get an AI summary, a practice's fees, services and practitioners are often described before a patient reaches its site.
- Google removed AI Overviews from some medical queries in January 2026 after an investigation found misleading health summaries, which shows how closely accuracy is watched in health.
- The facts worth publishing are practical ones: fees, rebates, telehealth, hours, referral rules, practitioner registration and languages spoken.
- Ahpra's advertising rules still apply to what you publish for AI: no testimonials about clinical care, no misleading claims and clear terms on any offer.
In this article
Patients are ready for AI in care
Patients across Asia-Pacific are broadly comfortable with AI in healthcare, and they expect providers to be easy to reach. Bain & Company surveyed 6,300 consumers in nine markets, including Australia, India and Singapore, for its Front Line of Healthcare report. Nearly three in four were comfortable with at least one AI-enabled healthcare application (fieldwork dates not stated).1 Bain published it in June 2026.
In the same survey, 71% expected doctors to be more responsive through channels such as phone, WhatsApp or email (fieldwork not stated).1 Bain does not break out Australian figures, so read these as regional.
Comfort with AI in care is not the same as asking a chatbot which clinic to book. But it removes one barrier. A patient who trusts AI to help with a symptom checker is likely to ask it which practitioners nearby do a procedure, what it costs and whether they bulk bill.
Our view: for an Australian practice the question is no longer whether patients ask AI. It is whether the answer they get about your practice is accurate and permitted.
Health questions get AI answers more often than most
Medical searches are among the most likely to show an AI summary. Ahrefs studied 146 million desktop search results from September 2025 (geography not stated). AI Overviews appeared on 20.5% of results overall, 44.1% of medical queries in its "your money or your life" group and 57.9% of question-style queries.2
That means a patient's first reading of your services, fees or practitioners is often a summary written by an engine. If your site does not state a fact clearly, the engine fills the gap from a directory, an old listing or a competitor's page.
Accuracy in health is watched closely. In January 2026 Google removed AI Overviews from some medical queries after a newspaper investigation found misleading summaries, including on the normal range for liver blood tests.3 Removal on clinical questions does not remove the risk on practical ones, such as what you charge or whether you offer telehealth.
What should a practice publish?
Publish the practical facts patients check before they book, as text on pages a crawler can reach. These are the facts AI answers repeat, and the ones most often wrong in directories.
Publish for AI answers, within the rules
Publish clearly
- Services and conditions you treat, described factually
- Practitioners, qualifications and registration status
- Fees, gap amounts, rebates and bulk-billing policy
- Telehealth, hours and after-hours arrangements
- Referral requirements and how to book
- Languages spoken and accessibility
Keep out of your own advertising
- Testimonials about clinical care
- Promised or implied outcomes
- Before-and-after images that set false expectations
- Offers without clear terms and conditions
- Urgency that encourages unnecessary treatment
- Comparisons you cannot support with evidence
Fees and rebates
State your standard consultation fees, the typical out-of-pocket gap after any rebate and your bulk-billing policy in a sentence each. Engines answer "how much does it cost" questions with whatever number they find. A clear, current figure on your own page gives them the right one to use.
Practitioners and registration
List each practitioner with qualifications, areas of practice and a note that registration can be checked on the public register. Keep it factual. "Special interest in" is fine where accurate; "the best" is not.
Access facts
Hours, after-hours arrangements, telehealth, whether a referral is needed, languages and accessibility are exactly the facts patients ask an assistant to filter by. Each belongs in text on your site, not only in a booking widget or an image.
Stay inside Ahpra's advertising rules
Everything you publish to help AI answers is still advertising a regulated health service. Ahpra's summary of the requirements, last reviewed in September 2025, says advertising must not be false, misleading or deceptive, and must not use testimonials about clinical aspects of care.4 It must not create unreasonable expectations of benefit or encourage indiscriminate or unnecessary use of services.4 Gifts and discounts need clear terms.4
This shapes AI work in two ways. You cannot quote patient praise of clinical outcomes on pages you control, even if engines would cite it. And the factual, plain style the rules require is also the style engines quote most easily.
Third-party reviews on independent platforms you do not control are a different matter. Engines will read them. Your job is to make sure your own pages state the facts correctly, so a summary drawn from reviews is anchored to something accurate.
Our view: Ahpra's rules make health practices better sources for AI, not worse. Stating fees, access and qualifications plainly, without superlatives, is what a careful engine wants to cite.
Keep every listing saying the same thing
Engines compare sources, and disagreement is where errors start. Your website, your Google Business Profile, booking platforms, health directories and your professional association listing should all state the same fees policy, hours, services and practitioners.
Make one person responsible for the list of places your practice appears, and update them together whenever something changes: a new practitioner, a closed day, a change to bulk billing. A practice that changes its hours on its website but not on two directories gives an engine three answers to choose from, and the engine may choose the oldest.
Retire what is no longer true. Remove departed practitioners' profiles rather than leaving them live, and redirect old service pages to current ones.
How to check what AI says about your practice
Ask the questions patients ask, on each engine, and compare the answers with what is true. Use treatment and cost questions, "which clinics nearby offer" questions, practitioner questions and comparisons. Record what each answer says about your fees, services, practitioners and access, and which sources it cites.
When an answer is wrong, trace it to the source before fixing anything. Our step-by-step fix for wrong AI answers covers tracing and correction. Our guide for clinics whose patients ask AI first covers the prompt set in more detail, and our page for healthcare providers describes how we draft prompts for a practice's specialties and countries.
Sources
- Bain & Company, Front Line of Healthcare APAC 2026: 6,300 consumers in nine markets (Jun 2026); fieldwork not stated
- Ahrefs, what triggers AI Overviews: 146M desktop SERPs, Sep 2025; geography not stated (Nov 2025)
- Euronews, Google removes some health-related questions from its AI Overviews (Jan 2026)
- Ahpra, summary of the advertising requirements (page reviewed Sep 2025)


