ChatGPT Ads for hospitals and clinics under the NMC advertising guidelines (2026)

Updated 8 October 2026 to the National Medical Commission's Guidelines on Ethical Advertising and Public Communication by Hospitals/Medical Institutions and Registered Medical Practitioners (public notice No. R-13014/01/2024-Ethics, dated 6 October 2026, in force immediately). Paragraph numbers in this article refer to those guidelines. General information, not legal advice.

This guide was first written around Chapter 6 of the 2002 Regulations; it now reflects the October 2026 guidelines. For the full rules, see the complete guide to the NMC advertising guidelines 2026. For AI-generated content outside paid ads (blogs, videos, social posts), see AI content in healthcare marketing and para 7.2.

What the October 2026 guidelines say about ads on AI platforms

The guidelines are titled "Guidelines on Ethical Advertising and Public Communication by Hospitals/Medical Institutions and Registered Medical Practitioners" (public notice No. R-13014/01/2024-Ethics). They are read with the IMC (Professional Conduct, Etiquette and Ethics) Regulations 2002, which remain the operative code of conduct, and with the NMC Act 2019 (paras 2.1(f) and 4.2). For hospitals, the Clinical Establishments Act or the relevant State Act prevails in an irreconcilable conflict (para 4.3).

Para 3.2 defines advertisement as any communication that directly or indirectly solicits or promotes. It lists digital, social-media, sponsored, targeted, influencer-led, search-engine, platform-based, audio-visual and AI-generated or AI-assisted promotional communication. Para 3.5 names AI-generated promotion and online healthcare platforms among the media of advertisement. A sponsored answer in an AI assistant sits squarely inside both definitions.

What an institution may say is set by para 8.3: factual, objective and verifiable information such as name, location, contact details, departments, facilities, equipment, services, emergency services, accreditation status and charges. It may not include solicitation, inducements, superiority claims, cure guarantees or promotional portrayal of an individual registered medical practitioner. Para 8.3 adds that the frequency or medium of publication must not turn factual information into solicitation, which matters for a channel that can repeat an ad across thousands of conversations.

Why conversational placements need a wider review

A search or feed ad is a fixed unit: a reviewer reads the headline, description and sitelinks once, and that exact text is what every viewer sees until the campaign is edited.

A sponsored response in an AI assistant behaves differently in three ways. The platform can rephrase, summarise or truncate the submitted copy. The ad often sits inside a multi-turn conversation, where a follow-up question ("is Dr. [Name] the best for this in Gurgaon?") pulls toward exactly the comparative framing the guidelines rule out. And any structured data the brand supplies (FAQ schema, Person schema, a knowledge-base feed) can be drawn into the follow-up answer.

Para 4.4 says publishing through an agency or platform does not absolve the institution of content it authorised, commissioned, sponsored, adopted or knowingly permitted. You cannot control a model's paraphrase, but you do control what it has to work with. So the review covers the ad unit, the landing page it routes to and the structured data on that page as one unit.

What clean copy looks like inside an AI-assistant ad

A simple test: remove every claim that para 8.1 or para 8.3 would rule out, and what remains should still answer the user's question. Factual detail does that job better than a superlative. For the phrase-level swaps, see the glossary entry on NMC-clean ad copy.

Both examples leave out the doctor's personal success rate, "best in the city" framing and "guaranteed" language. They also use "Learn" rather than "Book now", so that the ad informs instead of soliciting.

What happens if an ad is challenged

For registered medical practitioners, para 10.1 sets out graded action that "may be contemplated" by the State Medical Council: a warning with mandatory ethics training for a first violation, censure with a monetary penalty for a second, suspension of three to six months for a third, suspension of six to twelve months for serious violations (including inducements and digital mass solicitation), and removal from the register for one to three years for repeated violations. A show-cause notice comes first, followed by a reasoned order and a right of appeal (paras 10.2, 10.3 and 11). Hospitals and medical institutions are dealt with under the Clinical Establishments Act or the relevant State Act (para 10.1, Explanation). A complaint can also be made to ASCI about the advertisement itself.

Documentation helps in every route. A dated record of what was submitted, what was removed before launch and who approved it lets an institution show what it authorised (para 4.4). That is why ICG keeps the checklist as a permanent campaign record.