Is ChatGPT Ads an unregulated channel for fertility clinics?

Is ChatGPT Ads actually an unregulated channel for fertility marketing?

No, and the assumption is worth dealing with first because fertility marketing has a long history of people hoping a new channel has not been thought about yet.

OpenAI publishes an ad policy document covering three separate things: which advertisers may participate, what ads may contain, and where ads may be placed. It carries a last-updated date, which at the time of writing is 10 September 2026, and it has been revised repeatedly through the year.

Behind the policy sits a review process operating at three levels. Advertisers are assessed at signup for legitimacy and eligibility. Ad creative and landing pages are reviewed together when an ad is uploaded, covering title, copy, media and destination. Placement is checked as a final step. OpenAI says most of this runs through automated systems with human oversight, with escalation based on severity, confidence or risk.

So the position is closer to the opposite of unregulated. It is a young channel with a written rulebook, an enforcement process and a publication date, which is more than can be said for some established surfaces.

Why does a fertility clinic fail before the copy is written?

On two independent grounds, either of which would be enough on its own.

The first is the category. OpenAI restricts ads for regulated medical products and services, and lists what it may nevertheless approve from approved advertisers, case by case, in the United States: consumer medical devices, dental services, dietary supplements, disease awareness campaigns, health insurance, hospitals and urgent care, medical testing and diagnostic services, minimally invasive cosmetic procedures, and vision. Fertility treatment is not on that list.

The second is more specific and it would survive the category being added. Among the things OpenAI expressly does not allow are ads for invasive or high-risk medical procedures. Egg retrieval is an invasive procedure on any reading, and it sits at the centre of what an IVF clinic does.

There is a third layer that removes the question entirely for most of our clients. Outside the United States, ads for health services are generally prohibited, so a clinic in the UK, Australia, New Zealand or the UAE is not weighing a category decision at all.

How that reads for a clinic weighing the channel commercially is on ChatGPT Ads for fertility clinics.

Why is fertility excluded when dental is approved?

This is our reading rather than OpenAI's explanation, and we are labelling it as such because OpenAI publishes no reasoning.

It is a fair question, because fertility looks superficially like the approved service categories. Dental services and minimally invasive cosmetic procedures are both procedural, both delivered in a clinic, both elective for many patients. Fertility treatment is procedural too.

The word doing the work is invasive. The two approved service categories are bounded interventions that stay on the surface or just under it, and the cosmetic entry is qualified as minimally invasive rather than cosmetic generally. A treatment cycle involving stimulation, retrieval under sedation and transfer is a different order of procedure, and it falls inside a phrase the policy uses to exclude rather than one it uses to permit.

That reading suggests the exclusion is structural rather than an oversight waiting to be corrected. We would not plan a budget around it changing.

What would still govern a fertility ad if the category opened?

Three rulebooks, stacked, and only one of them is OpenAI's.

OpenAI's own baseline standards would apply first. Ads must be truthful and not misleading, including unfounded claims about capabilities, pricing, outcomes, affiliations or comparisons. Ads that exaggerate results or use false endorsements are prohibited regardless of category. For a sector whose central marketing asset is a success rate, that is not a peripheral rule.

Google's separate fertility advertising restrictions would still apply to your other channels, and they are the reason most fertility ads get rejected today. We covered them in why IVF ads get rejected by Google.

And your own regulator governs the claim whatever the platform decides. Approval by an advertising system is not permission from HFEA, from an advertising code, or from whichever body oversees you. Platform rules and regulatory rules are assessed by different people against different tests, and passing one has never meant passing the other.

Why is the success rate the hardest thing for a clinic to advertise?

Because it is simultaneously the most persuasive number a clinic holds and the most heavily governed one.

OpenAI bars unfounded claims about outcomes and ads that exaggerate results. Your own regulator governs how a success rate may be presented, which measure it uses and what has to sit alongside it. And in the UK the regulator publishes its own standardised version of that number about your clinic, which means any figure you advertise is readable against a version you did not write.

That last point is the one clinics underestimate, and it does not depend on any advertising channel. We set out how the regulator's record about a clinic keeps moving in what ChatGPT draws on about a fertility clinic across a months-long decision, and the presentation rules themselves in whether a fertility clinic can advertise its success rates.

So even in a hypothetical where the category opens, the strongest claim a fertility clinic has is the one it can least freely make, and the copy would have to be built from something else.

What is actually open to a fertility clinic on these surfaces?

The unpaid half, which for this sector is the better half anyway.

Being named or cited in an assistant's answer requires no category approval, is not affected by the placement rules that keep ads out of personal health conversations, and does not stop when a budget stops. For a decision that runs over months rather than minutes, that persistence matters more than it would in a trade where the purchase happens the same afternoon.

The work that earns it is specific and we have set it out elsewhere: what a clinic can publish that the regulator has not already published, and how the regulator's own record about you behaves over a long decision. Neither requires anybody's permission.

That is not a consolation prize for a closed channel. It is the work that would have had to happen first regardless, because an ad pointing at a page that cannot substantiate its claims fails on both sides.

Where does a fertility clinic stand on ChatGPT Ads?

How would a fertility clinic know if this changed?

By watching the policy page rather than the commentary, and by treating undated claims as worthless.

OpenAI's ad policies page carries both the approvable category list and the date it was last revised. If fertility were ever added it would appear there first. The Ads Manager availability list is the separate country check, and it is not the same question, because a clinic can be in a country where the platform operates and still be in a category that cannot advertise.

Everything here was read on 20 September 2026 from a policy last updated on 10 September 2026. That gap is ten days, and the policy moved several times before it. If you are reading this materially later, check the source rather than the page.

One thing will not move with it. The rules your own regulator applies to what you may claim travel onto any surface, new or old, and no platform approval has ever been a defence to a regulator.