ASCI Chapter III — aesthetic and cosmetic ChatGPT Ads guidance

What the rule actually says

ASCI Chapter III is the section of the Advertising Standards Council of India's Code for Self-Regulation that governs advertisements touching health, personal care and aesthetic outcomes. It was updated in 2023 to sharpen three areas that matter most for aesthetic and cosmetic advertising: before/after claims, superiority claims, and testimonials — and ASCI's Consumer Complaints Council has continued to rule against violations across digital formats through 2026, not just legacy print and television placements. ICG treats the current published wording of the chapter, not a summarised version from a training-data snapshot, as the source document for every compliance review, because ASCI issues periodic clarifications and the exact phrasing of a permitted disclaimer or an acceptable substantiation standard can shift between updates.

On before/after claims, the chapter requires that any depiction or description of a visible change resulting from a treatment be substantiated with genuine, current, comparable evidence — same lighting, same angle, same time-relevant context — and prohibits implying a typical or guaranteed result from what may be a best-case outlier. Importantly for a text-first ad format like ChatGPT Ads, the rule doesn't only apply to photographs; descriptive language that implies a specific verifiable visible outcome is treated the same way an unsubstantiated photo pair would be.

On superiority claims, the chapter requires that any "best," "number one," "most advanced" or comparably absolute claim be backed by objective, dated, producible evidence, and that comparisons not misleadingly disparage an unnamed category of competitor by implication. This is a stricter standard than many advertisers assume — a claim doesn't need to name a specific rival to trigger scrutiny if it implies unearned superiority over the category as a whole.

On testimonials, the chapter permits genuine patient testimonials but requires that they reflect an actual, current experience, avoid implying a typical or guaranteed result for a new patient, and where relevant carry a results-vary disclaimer. Testimonials that quote or imply a specific visible outcome inherit the same substantiation burden as a direct before/after claim, which is the detail advertisers most commonly miss.

How it applies to conversational-ad copy specifically

A static Google search ad or a Meta feed post is a fixed, reviewable unit — the exact text a compliance reviewer signs off on is the exact text every viewer sees. A ChatGPT Ads sponsored response is different in a way that matters directly to Chapter III: the initial ad copy is fixed and reviewable, but the conversation frequently continues past it, with the user asking a follow-up question that pulls the AI toward elaborating on the original claim. If the opening ad line says "visible skin improvement in as little as two sessions" and the user's next message is "how visible, can you show me," the compliance risk now depends on what the connected landing page and any linked content say next — not just on the original ad text.

This is why ICG's compliance review for aesthetic and cosmetic ChatGPT Ads covers three layers rather than one: the sponsored-response copy itself, the citation text that appears if the AI references the brand in its own generated answer, and the landing page the click resolves to. A brand can pass a Chapter III review on the ad copy alone and still be exposed if the landing page it sends traffic to contains an unsubstantiated before/after gallery or an outcome-guarantee headline — because ASCI's complaint process, and NMC's parallel process for the practitioner, both look at the full user journey, not the isolated ad unit.

The auction mechanics compound this. ChatGPT Ads bids are pegged to conversation-stage rather than a fixed keyword, so the same aesthetic brand might appear in an early, low-cost "is this treatment safe" exchange and, separately, in a late-stage, high-cost "should I book with this clinic" exchange. Chapter III applies with equal force at both stages, but the practical risk is higher at the late stage, because that is where a user is most likely to ask a direct efficacy or outcome question the AI has to answer using whatever the brand has published — which is why ICG writes landing-page content for the booking-ready intent bucket with the same before/after and testimonial discipline as the ad copy itself, not a looser standard.

Attribution on conversation-completion events, rather than clicks, also means a non-compliant elaboration doesn't need to produce a click to matter — a user who reads a misleading AI-generated elaboration on a brand's claim and then abandons the conversation has still been exposed to non-compliant content the brand's own account made possible, even without a measurable ad interaction. ICG treats this as a reason to over-invest in guardrail language within the ad copy itself, since the brand cannot fully control what the model generates in response to an open-ended follow-up.

Common violations and how to avoid them

The most frequent Chapter III issue ICG catches in pre-existing aesthetic ad copy is unsubstantiated before/after language dressed as a neutral description rather than a claim. An anonymised example from a dermatology brand's draft copy read "see dramatic pigmentation clearance in just 3 sessions" — this reads as a factual timeline but is functionally a before/after claim with no substantiation file behind the specific number, and ASCI has upheld complaints against near-identical phrasing in other categories. The fix ICG applied was to shift the claim to a range grounded in what the clinic could actually document — "many patients notice visible improvement over a course of sessions, typically 4–8 depending on pigmentation depth" — which preserves the marketing message while removing the unsubstantiated specific-outcome guarantee.

The second most common issue is superiority language with no evidence file behind it. A hair transplant brand's draft opened with "India's most advanced FUE technique" — a superiority claim that would require dated, objective, producible evidence of category leadership to survive an ASCI complaint, evidence that essentially no single clinic in a fragmented category can actually produce without disputable methodology. ICG's standard fix is to replace the superiority claim with a specific, factual differentiator the clinic can actually substantiate — graft survival methodology, surgeon case volume, or equipment specification — rather than an unearned absolute.

The third recurring issue is a testimonial quote that smuggles in an outcome claim. A cosmetic surgery clinic's draft testimonial read "I lost 3 sizes and my confidence came back" attributed to a named patient — the specific numeric outcome inside a testimonial inherits the same substantiation burden as a direct before/after claim, and pairing it with an emotional confidence line doesn't exempt it. ICG's fix strips the specific numeric outcome from any quoted testimonial and keeps testimonial content focused on the experience — communication quality, comfort during the procedure, aftercare support — which is defensible without a substantiation file because it isn't an outcome claim at all.

A fourth issue specific to the conversational format: brands sometimes write ad copy that reads as neutral, unbiased information rather than promotional content, in an apparent attempt to blend into the AI's own answer style. This runs against the sponsored-content disclosure requirement in spirit even when the platform's own sponsored label is technically present, because the intent of the disclosure rule is that a consumer recognise promotional content as promotional. ICG writes aesthetic ad copy to read honestly as an offer — naming the clinic, naming the treatment, making a clear promotional ask — rather than mimicking the AI's own neutral, informational tone.

What "clean copy" looks like inside a ChatGPT Ad for this domain

Below are worked before/after pairs from ICG's aesthetic ChatGPT Ads practice, each anonymised and each representing a real compliance correction made before launch.

The rewrite removes the outcome guarantee, replaces the single-session claim with a realistic multi-session range, drops the unearned superiority line, and adds a results-vary qualifier without losing the promotional intent of the ad.

The rewrite drops both the unattributed testimonial paraphrase and the permanence guarantee, replaces them with a factual, verifiable process detail (the graft-survival protocol) and a realistic development timeline, and moves the call to action to an assessment rather than an implied commitment to a specific density outcome.

A third pattern ICG applies consistently across the aesthetic and cosmetic vertical: pricing and package language stays factual and range-based ("packages typically range from ₹X to ₹Y depending on area treated") rather than framed as a discount urgency claim, because Chapter III scrutiny extends to misleading pricing presentation as well as outcome claims, and an aesthetic patient researching cost inside a ChatGPT conversation responds better to an honest range than to a limited-time framing that reads as manufactured urgency in a medical context.

ICG's compliance checklist before every ChatGPT Ad push

Every aesthetic and cosmetic ChatGPT Ads bucket ICG builds passes through the same written checklist before it goes live, run against the current published wording of Chapter III rather than a cached summary:

• No before/after claim — descriptive or photographic — without a documented, current, comparable substantiation file on record.
• No superiority language ("best," "number one," "most advanced," "leading") without objective, dated, producible evidence; specific factual differentiators used instead.
• Any testimonial reflects a genuine, current patient experience, avoids a specific quantified outcome claim, and carries a results-vary qualifier where relevant.
• Recovery, duration and efficacy language stated as a realistic range, never a best-case number presented as typical.
• No implied outcome guarantee anywhere in the ad copy, citation text, or connected landing page — checked as one journey, not three isolated units.
• Sponsored-content intent is honest in tone, not written to mimic the AI's own neutral informational voice.
• Pricing presented as a factual range, not framed with artificial urgency or discount pressure.
• DPDP 2023 consent language present and unambiguous on every connected lead form, with no pre-ticked marketing consent.

This checklist runs at three points in the account lifecycle: before initial launch, whenever ASCI or NMC issues a public update or clarification, and whenever an existing ad's connected landing page is edited — because a landing-page change can silently break compliance on an ad that was fully compliant at launch. ICG logs the checklist result for every bucket, not just the launch date, so a clinic has a documented compliance trail if a complaint or an internal audit ever asks for one.

For clinics that write some or all of their own ad copy, ICG runs the identical checklist against externally drafted copy before agreeing to push it live inside a managed ChatGPT Ads account — a clinic's in-house draft gets the same scrutiny as copy ICG wrote itself, because the compliance exposure sits with the campaign and the practitioner's registration, not with who typed the words.

What happens if you're audited

An ASCI complaint typically begins with a consumer, competitor or the council's own monitoring flagging an advertisement; the Consumer Complaints Council reviews it against the current code and can direct the advertiser to modify or withdraw the ad, usually within a defined response window. Persistent or serious non-compliance can be escalated further, including to sectoral regulators for digital media, and for a healthcare advertiser it can also trigger a parallel NMC Section 6 inquiry against the named practitioner independent of the ASCI process — the two tracks run separately and a clean outcome on one doesn't resolve exposure on the other.

Because a ChatGPT Ads sponsored response and its connected landing page are both time-stamped and archivable, an audit response is materially easier when the compliance review itself was documented at launch rather than reconstructed after a complaint arrives. ICG keeps the checklist result, the reviewed copy version, and the date of review for every bucket in an aesthetic or cosmetic account, so a clinic facing a complaint can produce its substantiation position — what evidence backed a given claim, and when it was reviewed — rather than starting that documentation from zero under time pressure.

None of ICG's aesthetic or cosmetic ChatGPT Ads clients has had an ad upheld against under Chapter III to date, which ICG attributes directly to reviewing every bucket before launch rather than reactively after a complaint — a discipline that costs nothing extra beyond the standard tier fee, since compliance review is built into every ICG ChatGPT Ads engagement rather than sold as a separate audit product.