Crawl Nine

Which businesses can advertise on ChatGPT?

Published · Last updated

Sources last checked . This page describes a product that changes often. Every claim is linked to the page it came from, so you can check any of it yourself.

Which businesses can advertise on ChatGPT?

Two things decide it. Where your business is registered, which is now open in 63 countries. And what trade you are in, which is not open to everyone: OpenAI limits advertising to a set of consumer categories while the product is being tested, its own word is “primarily”, and handles legal, health and financial services separately, with different rules inside and outside America.

The trade gate catches far more businesses than the country gate does, and almost nobody writing about ChatGPT advertising mentions it. This page states both, with OpenAI’s own wording, so you can tell within a minute which side of the line your business is on.

Everything below comes from OpenAI’s ad policies and advertiser documentation, read on 4 October 2026. OpenAI’s policy page carries a version number, and it has moved six times since March.

Which trades can advertise on ChatGPT today?

OpenAI states that during the initial test period ads are “primarily limited to consumer verticals such as lifestyle and household goods, local services, travel and experiences, and digital products or education.” OpenAI does not define “local services”; on any ordinary reading it covers the trades, plumbers, electricians, heating engineers, roofers.

If you are a trade, a shop, a hotel or somebody selling a course, you are on the right side of the category line. That is the gate this section is about, it is not approval. Every advertiser still goes through identity checks and every ad still gets reviewed, and OpenAI hedges its own wording twice over: “During the initial test period, ads are primarily limited to” those verticals.

OpenAI adds that these “categories may expand over time”, and separately that everything else is “disallowed at launch”, naming dating or sexual content, health claims, alcohol and drugs, healthcare, financial or legal services, gambling and political content among the exclusions. That list “may evolve as the program develops.”

So the honest summary is: the category is open for consumer trades, closed for the excluded ones, and country-dependent for legal, health and financial services, which is the next section. Being in an open category is necessary rather than sufficient. OpenAI’s policy also says it “reserves the right to decline, restrict, or remove advertisers and to refuse to display ad content and website links for any reason”. In plain terms, OpenAI can turn a business away even when its trade is allowed.

Can a law firm advertise on ChatGPT?

In America, yes, if the firm is licensed where the ad runs. OpenAI states that legal ads are “permitted in the US only when the advertiser is licensed to practice law in the jurisdiction where the ad is shown”. Outside America they are not: “Ads for legal services outside of the US are currently prohibited.”

This changed on 31 August 2026 and it is a genuine reversal, not a clarification. OpenAI’s own changelog says why that version exists: “v1.5 (August 2026): Updated to reflect that legal services are permitted in the US.”

The permission is specific about what it covers, “services related to immigration, personal injury, legal claims, or document preparation”, and about what does not need it. Ads for “general legal education or media may be allowed where no legal services are offered”, with legal podcasts and LSAT courses given as OpenAI’s own examples.

One thing to hold in mind while reading the policy: an earlier section of the same page still lists legal services among categories approved case by case, so approval is not automatic even in America. Expect to show a licence.

Can a dentist, clinic or med spa advertise on ChatGPT?

In America, possibly. OpenAI states that in the US it “may allow ads from approved advertisers on a case-by-case basis” for a named list that includes “Dental services” and “Minimally invasive cosmetic procedures”. Outside America, almost certainly not: “Ads for health services outside the US are generally prohibited.”

Being named in that list matters. It is the difference between a category OpenAI has not thought about and one it has decided it can approve. The full list also covers consumer medical devices, dietary supplements, disease awareness campaigns that stay off public policy, health insurance, hospitals and urgent care, medical testing, and vision. That is all nine.

OpenAI’s help centre carries its own table for health ads, and it does not match the policy line for line. It has no row for dental services or for vision. It marks “Minimally invasive cosmetic procedures” as “Allowed” in the US, and marks “Hospitals and urgent care”, “Health insurance” and “Medical testing and diagnosis” as “Restricted to Select Approved Advertisers”. It also says “At this time we are only working with a small group of advertisers and are not accepting applications for approval in certain policy categories.” So a US dentist should confirm with OpenAI that dental ads are open before planning a campaign.

Approval is not a formality. OpenAI requires that eligible ads “must be lawful, factual, responsibly presented, and subject to appropriate review and compliance controls”, and says “Advertisers may be required to provide proof of licensure.”

Some things stay out regardless: “unsafe or unapproved medical products; unsupported treatment claims; experimental or investigational therapies; invasive or high-risk medical procedures”.

If you sell fitness equipment or menstrual products rather than treatment, you are in easier territory, OpenAI says general health and wellness products that make no medical claims may be permitted.

Where can you open an advertiser account?

In 63 countries. When we read OpenAI’s availability table on 4 October 2026, every country listed was marked available for self-service and none was marked coming soon. What matters is where the business is registered, not where its customers are.

Europe was the biggest block still missing, and it opened on 31 August 2026, alongside India, the Middle East and North Africa, which OpenAI announced the same day. It added an update to its own European announcement confirming self-service across all 31 European markets. The table has grown since. On 23 September 2026 OpenAI announced seven more markets in Asia: Indonesia, Malaysia, the Philippines, Singapore, Thailand, Vietnam and Taiwan. All seven are now on the table, and so are Israel, Saudi Arabia, Turkey and the United Arab Emirates.

OpenAI’s condition is about the legal entity: “the legal entity that will advertise and be billed must be based in a country listed as available”. So what decides it is where the business is registered and billed: one registered outside the list cannot buy self-service even if its customers are inside it. A new account may also be limited at first in where its ads can run: OpenAI says “Some new self-serve ad accounts can initially advertise only in their home country”, and lifts that limit automatically once identity verification is approved and the account has spent enough on campaigns in its home country.

OpenAI shows a relative “Updated” note on that table rather than a date, so it can change without anything visibly changing. Check it against the link below rather than trusting this page’s figure in six weeks.

Where will your ad never appear?

OpenAI keeps ads out of whole categories of conversation regardless of who is advertising. Its placement policy names nineteen kinds, and does not present them as the whole list. Among them: child safety, self-harm and suicide, weapons, terrorism, hate and harassment, political content, regulated goods such as gambling and tobacco, misinformation, privacy and graphic content.

There is a second list, and it is the more interesting one commercially. OpenAI also withholds ads from “conversations exhibiting vulnerable user–model interactions”, naming emotionally reliant contexts and mental and personal health among them.

The practical effect is that reach is not the same as audience size. If your customers tend to arrive through a conversation OpenAI classes as sensitive, your ad will simply not be there, and no budget changes that.

This is also part of why nobody can quote you a reliable reach figure for “ChatGPT users”. Ads only reach people on the Free and Go plans, OpenAI states that “Plus, Pro, Business, Enterprise, and Edu accounts will not have ads”, nor does any account it identifies as belonging to an under-18 see them, nor do these conversations.

If you cannot advertise, what can you do?

Get named inside the answer instead. That is not sold, and the advertising category rules do not apply to it. ChatGPT names businesses in its answers based on what it can find and read on the open web. A clinic in Britain that cannot buy an ad can still be the business ChatGPT recommends.

These are two separate systems and OpenAI is explicit that they do not touch: advertising “does not influence the answers ChatGPT provides”. That cuts both ways. You cannot buy your way into the answer, and being barred from advertising does not keep you out of it.

The work is different: making your site readable to the assistants that cannot run JavaScript, answering the questions people actually ask, and being findable in the search index ChatGPT draws on. That is what Discover is.

The first step costs nothing and needs no email, check whether the assistants can read your site.

Sources

All fetched 4 October 2026.