ChatGPT Ads policies: what is allowed, restricted or prohibited

The principle that shapes everything else

Before the categories, a principle OpenAI states plainly, and which explains how strict the rest is: ads do not influence ChatGPT's answers. Ads run on separate systems from the chat model, and advertisers have no ability to shape, rank, or alter ChatGPT's responses.

Ads appear below ChatGPT responses, and only users on the Free and Go plans may see them. Plus, Pro, Business, Enterprise and Edu accounts do not have ads.

That strict separation between answer and advertisement is why the truthfulness rules matter so much: a misleading ad inside a conversational environment, where the user extends significant trust to the context, is a different order of problem than on a classic results page.

The categories allowed during the test period

The starting point is restrictive, and it needs reading correctly: this is not a list of prohibitions inside an otherwise open space, it is a list of permissions, with everything else closed. During the initial test period, ads are primarily limited to consumer verticals:

• Lifestyle and household goods
• Local services
• Travel and experiences
• Digital products or education

All other categories are disallowed at launch. If your business doesn't obviously map to one of those four families, the question isn't "how do I word my ad" but "am I eligible at all".

The three case-by-case categories

Three industries get special handling: OpenAI says it may approve ads from approved advertisers within the financial services, healthcare and medicine, and legal services categories. These categories are being rolled out gradually, with approvals reviewed manually on a case-by-case basis.

Read that this way if you're in one of those industries: access is neither automatic nor ruled out on principle, but it doesn't plan like a simple campaign activation. Budget for a delay and a submission, and don't commit a date to leadership on the strength of finished creative alone.

One further constraint: advertisers whose primary business model or offerings fall within restricted or prohibited categories (healthcare or weapons, for example) may be ineligible to advertise on the platform, regardless of the campaign proposed. That filter applies at account verification time, a point covered in our guide to creating a ChatGPT Ads account.

The prohibited categories

Alongside ads that violate OpenAI's usage policies, prohibitions cover sensitive or regulated areas:

On gambling, the wording is broad: ads promoting products or services that involve wagering money on games of chance, skill, or uncertain outcomes are currently disallowed, including casino or betting promotions, online sports betting, lotteries, and poker or gaming platforms offering cash prizes.

For a merchant, these categories need checking catalog line by catalog line before connecting a whole feed: see our guide to ChatGPT Ads for ecommerce.

On political content, ads that advocate for or against political actors, elections, public policy, or contested social issues are currently disallowed.

Housing and employment: the nuance that matters

This is what OpenAI clarified in August 2026, and it's worth reading precisely because the rule is not a blanket ban:

• Ads for individual job listings or individual housing rentals or sales are prohibited.
• Advertisers may link to platforms that have job or house listings, as long as the landing page or ad creative does not reference a specific listing.

Concretely: a jobs marketplace can promote itself as a service, it cannot promote "developer role at X in Manchester". The distinction is about the granularity of what's promoted, not the industry.

Credit, debt and alternative investments

A specific and explicit rule: ads for credit repair, debt settlement, debt assistance, or alternative investments such as bullion are not allowed.

This one regularly catches advertisers who mentally file themselves under coaching or advisory rather than financial services. The rule reads on the nature of the offering, not on how the business describes itself.

Content requirements: truthfulness above all

Regardless of category, the policies require that ads be truthful and not misleading. Specifically targeted:

• Unfounded claims about capabilities, pricing, outcomes, affiliations, or comparisons with other products or companies.
• Ads that exaggerate results.
• False endorsements.
• More broadly, any ad that deceives users, regardless of category.

Two practical consequences for copywriting. First, an unsupported competitive comparison is grounds for rejection even inside an allowed category. Second, if your usual pitch rests on a performance number, have your evidence ready before it goes into an ad.

Landing page requirements

The destination page is inside the compliance perimeter just as much as the creative. Two documented requirements:

• Landing pages and destinations must accurately represent the advertiser, product, or service described in the ad and may not mislead users through deceptive practices.
• All landing pages must use valid links and must not block OpenAI user agents, OAI-AdsBot and OAI-SearchBot.

OAI-AdsBot is required for ChatGPT Ads landing page validation and review. Common causes of disapproval include exactly this: a landing page that blocks OpenAI user agents or IP addresses, and ad copy or landing pages that do not align with the ad policies.

If your infrastructure runs a WAF, a CDN, bot mitigation or aggressive rate limiting, test before you launch with our ChatGPT Ads landing page checker.

Licenses, registrations and regulated professions

A simple rule with heavy consequences: advertisers promoting products or services that require professional registrations, licenses, certifications, or similar authorizations must maintain those credentials and comply with applicable laws and regulations.

This covers far more trades than people expect: real estate agents, brokers, healthcare professions, accredited training, transport, private security. Assume these may be requested during review rather than hunting for them under time pressure.

Prohibited data and personalization

On the data side, the policies define Prohibited Data that cannot be used for advertising purposes. It includes data relating to, or collected from, services directed to children, as well as data relating to sensitive information, with categories such as sexual orientation, race or ethnic origin, religious beliefs, financial distress, political opinions or affiliation and other sensitive categories.

European context reminder: personalized ads are not initially available in the European Economic Area or Switzerland. On the user side, OpenAI documents the ability to control ad personalization, turn it off, and clear the data used for ads at any time, with conversations kept private from advertisers.

Review, rejection and appeals

The contractual frame is broad: OpenAI may review, reject, remove, or require modification of any Ad at any time, for any or no reason, including non-compliance with law, the Advertising Terms, the OpenAI Policies, or applicable technical specifications.

If an ad is rejected, two documented routes:

1. Edit the ad and resubmit it for review.
2. Request another review if you believe the ad was rejected in error.

The common causes to check first: creative or landing page that doesn't align with the ad policies, a landing page blocking OpenAI user agents or IP addresses, an ad that is misleading or inconsistent, or disallowed content.

Before you launch: the six-point compliance review

1. Is your category on the allowed list (lifestyle and household goods, local services, travel and experiences, digital products or education)? If not, does it fall into one of the three case-by-case categories?
2. Does your primary business model fall into a restricted or prohibited category, which can make the account ineligible?
3. Is every claim in your ad supportable: capabilities, pricing, outcomes, affiliations, comparisons?
4. Does the landing page accurately represent what the ad promises, on a valid link?
5. Can OAI-AdsBot reach the page, counting robots.txt, WAF, CDN, bot mitigation, authentication and rate limiting?
6. Are your licenses and registrations current and available if review asks for them?

And a method rule that applies to this whole page: the policies change. August 10, 2026 is the latest known update as we write, but check the current version on OpenAI's official pages before launching in a sensitive industry. This is the one section of our cluster where we explicitly ask you not to take our word for it.