Which Categories Are Banned From ChatGPT Ads

Last reviewed 25 August 2026.

Before you plan a budget, check whether you are allowed to spend it. Ads inside ChatGPT are blocked from appearing near content relating to health, mental health, politics, gambling, legal services and financial services.

That list removes a striking amount of the digital advertising economy. Personal injury law, insurance, mortgages, wealth management, addiction treatment, therapy, supplements, sports betting: some of the highest cost per click categories in Google Ads are simply not addressable here.

Why the restrictions exist

The reasoning is straightforward once you consider the format. An ad inside a conversation sits directly beneath an answer the user is treating as advice. When the subject is a health symptom, a legal problem, or a financial decision, the proximity between advice and advertisement becomes a genuine harm risk rather than a branding concern. A sponsored card next to a mental health answer is a different object from a banner on a website.

There is also a regulatory dimension. Advertising in these categories is heavily regulated across jurisdictions, and building compliant systems for them is slow work. Blocking first and opening later is the sensible sequence.

The trap of adjacency

The restriction is on the conversation, not only on the advertiser. This catches people out. A perfectly ordinary ecommerce brand can find impressions suppressed when its contextual match lands in a conversation that drifts into a health topic. If you sell athletic supplements, running shoes, or food products, expect a share of your intended context to be unavailable.

This means restricted category exposure is a spectrum, not a binary. Check where your product conversations actually go, not just what industry code you file under.

What to do if you are blocked

You have one route into this channel and it is not paid. It is getting cited organically inside AI answers when the assistant responds to a buying question in your category.

The mechanics differ from classical search optimisation. Assistants synthesise from sources they consider authoritative and corroborated, which puts weight on being named consistently across independent third party sources rather than on ranking a single page. Structured, clearly attributed content that directly answers specific questions performs better than long form content optimised for keyword coverage.

For regulated categories this is arguably a better position anyway. Organic citation compounds and does not disappear the day you stop paying, and it does not carry the compliance exposure that paid advertising in these verticals does.

Will the restrictions lift?

Probably, partially, and not soon. Every major ad platform has followed the same path of opening restricted categories gradually with certification requirements attached, as Google did with addiction treatment and financial services. Expect certification and verification programmes rather than a blanket opening.

Planning assumption: if you are in a restricted category, build organic visibility now and treat eventual paid access as upside rather than strategy.

Checking your own position

Three questions. Does your product sit inside one of the six named categories. Do your buyer conversations routinely touch one of them even if your product does not. And can your product be explained without visual demonstration, since conversational formats suit products that are described rather than shown.

If the answer to the first two is no, you are eligible and the reference page covers what to do next. If either is yes, organic visibility is your route and we should talk about that instead.

Leave a Comment

Your email address will not be published. Required fields are marked *