ChatGPT ads policy: restricted categories and the rules for healthcare and finance advertisers
Updated 7 min read
In short
ChatGPT ads policy, as of October 2026, allows most consumer categories such as lifestyle goods, local services, travel, and education, while treating financial, health, and legal services as restricted, US‑only categories that require case‑by‑case approval and, in some cases, proof of licensure. Adult content, alcohol and tobacco, gambling, political content, recreational drugs, scams, sensitive-event exploitation, and unsubstantiated wellness products are banned. All ads must be truthful, non‑misleading, brand‑safe, and avoid sensitive user contexts per OpenAI’s documented standards.
Key takeaways
As of October 2026, ChatGPT ads are broadly open to consumer categories such as lifestyle, local services, travel, and digital education products.
Financial, health, and legal services are restricted US-only categories that require case-by-case approval and may require proof of licensure.
Adult content, alcohol and tobacco, gambling, political content, recreational drugs, scams, and unsubstantiated wellness products are disallowed categories.
All advertisers must meet baseline standards on truthfulness, non-discrimination, destination integrity, and honest business practices across creatives and landing pages.
Ads are excluded from sensitive and brand-unsafe contexts, and OpenAI uses multi-level review and ongoing monitoring with the right to restrict or remove ads.
As of October 2026, OpenAI’s ChatGPT ads policy centers on consumer-friendly categories, with regulated verticals like finance, health, and legal allowed only under tight US-only restrictions. For status and country eligibility, see the ChatGPT ads overview, and for setup steps and account review sequencing see the how-to guide.
Categories that can advertise freely today
Per OpenAI’s documented ad policies as of September 2026, launch inventory is concentrated in mainstream consumer categories. These include lifestyle and household goods, local services, travel and experiences, and digital products and education. These categories are treated as generally allowed, subject to the baseline content and advertiser standards described later in this guide.
Within these allowed areas, advertisers still need to avoid disallowed products and claims. For example, a travel advertiser can promote lodging and experiences, but not gambling if the property is a casino with a gambling focus. A wellness brand can promote general fitness or education, but not unsubstantiated diet pills or detox programs, which are explicitly disallowed.
Being in an allowed category does not guarantee that an ad will run in every ChatGPT context. OpenAI’s policies state that placement rules keep ads away from sensitive and brand-unsafe conversations regardless of advertiser type, so even fully compliant consumer brands will see some inventory excluded.
Restricted categories and what advertisers must prove
OpenAI treats financial services, health services, and legal services as restricted categories as of October 2026. These are generally allowed only in the United States, on a case-by-case basis, and may require proof of licensure or professional credentials according to OpenAI’s ad policies.
| Restricted category | Examples listed by OpenAI | Key conditions (as of Oct 2026) |
|---|---|---|
| Financial services | Auto loans and leasing, credit cards, credit monitoring, deposit accounts, financial planning, insurance, investment services and brokerages, mortgages, personal loans, payment services | US-only, case-by-case approval; licensure proof may be required; credit repair, debt settlement, debt assistance, and alternative investments such as bullion are not allowed; financial services ads outside the US are generally prohibited. |
| Health services | Consumer medical devices, wearables, health apps, dental services, dietary supplements, disease awareness (non-policy), health insurance, hospitals and urgent care, medical testing and diagnostics, minimally invasive cosmetic procedures, vision products and services | US-only, case-by-case approval; unsafe or unapproved products, unsupported treatment claims, experimental therapies, invasive or high-risk procedures, and body-image-exploiting products are not allowed. |
| Legal services | Immigration, personal injury, claims, document preparation, and similar services; legal education and media without services may be allowed | Permitted in the US where the advertiser is licensed in the jurisdiction where the ad shows; legal services outside the US are currently prohibited. Legal services became permitted in the US in the August 2026 policy revision. |
For these restricted categories, OpenAI’s documentation indicates that advertisers may need to demonstrate truthful identity, professional licensure where applicable, and compliant business practices. Financial and health advertisers must avoid subcategories that OpenAI explicitly disallows, such as credit repair and experimental therapies. Legal advertisers must be licensed in the jurisdictions where their ads appear.
A common misconception is that US approval in a restricted vertical unlocks global delivery. OpenAI’s policies state that financial, health, and legal services permissions are US-only, and that these categories are generally prohibited outside the US, so advertisers planning multi-country campaigns need to plan separate strategies for non-US markets.
Categories that are banned outright
OpenAI’s ad policies list several disallowed categories that cannot advertise on ChatGPT as of October 2026. These are banned regardless of the advertiser’s credentials or geography.
| Disallowed category | Examples and notes |
|---|---|
| Adult content and dating | Sexual content, adult entertainment, and dating services are disallowed. |
| Alcohol and tobacco | Alcohol over 0.5% ABV, cigarettes, vaping products, and nicotine products are disallowed. |
| Counterfeit goods | Products infringing trademarks or intellectual property are disallowed. |
| Gambling | Gambling services are disallowed; casino lodging or entertainment without a gambling focus may be permitted, as may non-cash digital games. |
| Graphic sexual or violent content | Ads with graphic sexual or violent imagery or descriptions are disallowed. |
| Individual housing and job listings | Individual property or job postings are disallowed; platform or marketplace ads without specific listings may be allowed. |
| Political content | Ads about elections, public policy advocacy, or contested social issues are disallowed. |
| Recreational drugs | Cannabis and THC products are disallowed; non-intoxicating hemp or CBD topicals may be permitted. |
| Scams and fraud | Deceptive or fraudulent schemes are disallowed. |
| Sensitive topics and events | Exploitation of sensitive topics or events is disallowed. |
| Unsubstantiated wellness claims | Diet pills, detox programs, and coached health claims that are unsubstantiated are disallowed. |
Being disallowed today does not mean a category is permanently excluded. OpenAI’s changelog shows that legal services, which were initially not permitted, opened as a restricted US-only category in August 2026, so a current “no” can change as policies evolve.
Baseline standards that apply to every ad
OpenAI’s policies define baseline standards that apply to all ads, regardless of category. Ads must be truthful and not misleading, which includes avoiding unfounded capability, pricing, or outcome claims, false endorsements, and exaggerated results. This applies both to the ad copy and to any claims made on the landing page.
Ads may not use obscene, vulgar, or shocking language or imagery, including in product names. They must avoid discriminatory, defamatory, or exclusionary content, including slurs or attacks based on protected characteristics. These standards are in addition to OpenAI’s general usage policies, which also apply to advertising content.
OpenAI requires that ads do not imitate ChatGPT’s interface, functionality, or voice in a way that could make users believe the ad is part of the product. Creative that mimics system messages or the chat UI in a confusing way can be rejected under this rule. Advertisers need to make the promotional nature of the content clear.
There is also an end-to-end consistency requirement. OpenAI’s documentation states that approved creatives may not link to destinations that introduce disallowed content, such as a food delivery ad that sends users to a page focused on alcohol delivery. If the landing page contains disallowed products or claims, the ad is ineligible even if the creative itself appears compliant.
Advertiser-level rules also apply across all categories. Advertisers must present truthful identity and affiliations, use only authorized intellectual property, maintain destination integrity, avoid platform abuse, and operate safe and honest business practices. Where professional credentials are required, such as in restricted financial, health, or legal services, those credentials must be maintained.
Contexts where ChatGPT will not show ads
OpenAI’s ad policies describe placement exclusions that apply regardless of advertiser category. Ads do not appear near sensitive user contexts, such as mental and personal health conversations, emotionally reliant contexts, or other sensitive user journeys. This means that even compliant health or wellness advertisers will not have their ads shown in certain 1:1 support-like conversations.
Ads are also excluded from brand-unsafe contexts. OpenAI lists child safety, dangerous activities, debated social content, fraud and deception, graphic violence, hate and harassment, illicit content, misinformation, political content, regulated goods, suicide and self-harm, terrorism, weapons, and similar topics as contexts where ads will not be placed. This placement policy corrects the misconception that a compliant category guarantees visibility in all ChatGPT conversations.
How review and enforcement work
OpenAI describes a three-level review system for ads. First, advertiser review covers verification, account quality, and risk signals. Second, creative and landing page review evaluates the ad assets and the destination page, and any ad whose landing page cannot be reviewed is ineligible to run. Third, placement review ensures that ads are matched only to suitable contexts.
Review is mostly automated, using language models and classifiers, with human oversight. Restricted categories such as financial, health, and legal services may require enhanced verification or manual review before approval. This corrects the misconception that only the ad creative is reviewed; OpenAI explicitly reviews advertiser identity, creative, media, and landing pages.
Enforcement actions range from ad rejection or removal, to limited delivery, to required edits, and up to advertiser suspension. OpenAI’s documentation notes that monitoring continues after approval, and users can report ads directly within the product. If an ad or advertiser later violates policy, OpenAI can adjust enforcement accordingly.
OpenAI also states that it reserves the right to decline, restrict, or remove advertisers and ads for any reason, including conflicts with its advertising principles, business interests, or competitive position. This gives OpenAI broad discretion beyond the specific category rules and content standards.
How quickly ChatGPT ad rules are changing
OpenAI publishes a changelog for its ad policies, which shows frequent updates in 2026. Version 1.0 launched in March 2026, followed by v1.1 in April (refining regulated-advice placement so that medical, legal, and financial advice contexts were no longer categorically blocked), v1.2 in May (adding a standards implementation section), v1.3 in July (clarifying advertiser policies and financial and health categories), v1.4 in August (clarifying housing and jobs), v1.5 in August (permitting legal services in the US), and v1.6 in September (clarifying OpenAI’s right to decline ads).
This cadence shows that a category’s status can change within a few months. For example, legal services moved from not being permitted to being allowed as a restricted US-only category in August 2026. A “no” answer for a given category today does not guarantee a permanent ban, so evaluators should check the live policy page regularly, especially if they operate in regulated verticals.
Because OpenAI can also refine placement rules, advertisers should expect that where and how often their ads can appear may change even if their category status does not. For deeper guidance on which allowed categories tend to perform well in conversational environments, see this overview of strong ChatGPT ad use cases.
Frequently asked questions
Per OpenAI’s ad policies as of October 2026, financial services advertising is generally limited to the United States. The policy states that financial services ads outside the US are generally prohibited, and that listed financial categories such as auto loans, credit cards, and investment services are reviewed case by case in the US only. Non-US financial brands should expect not to be eligible until the policy explicitly changes.
Sources
This resource is for educational purposes. Ad platforms change their policies, features, and pricing frequently; verify details against the platform's current documentation before acting on them.
This resource is for educational purposes and is not legal or compliance advice. Healthcare advertisers should confirm their obligations under HIPAA and other applicable regulations with qualified counsel.
This resource is for educational purposes and is not financial, legal, or compliance advice. Financial-services advertisers should confirm their regulatory obligations with qualified counsel before acting.
