OpenAI has opened advertising in Europe

But who is allowed to advertise, where may ads appear, and how does it work? Read on to learn more.

OpenAI has a fairly detailed Ads Policy, most recently updated on August 10, 2026.

In brief

There are two distinct levels of restrictions:

1. The types of ads that may be purchased.

2. The conversations alongside which ads may appear at all.

🟢 Categories that are generally eligible

In the current phase, ChatGPT Ads primarily focuses on:

• Consumer goods and household products

• Lifestyle

• Local services

• Travel and experiences

• Digital products

• Education

In some cases, OpenAI may also approve advertisers in finance, health and medicine, and legal services. These are reviewed individually and are not generally open categories.

🔴 Prohibited or heavily restricted categories

The most important include:

• Politics – including candidates, parties, elections, referendums, and political lobbying

• Alcohol and tobacco or nicotine

• Gambling – including casinos, betting, lotteries, and poker

• Adult content and dating

• Sexual products and services

• Cannabis, psychedelics, and other drugs

• Weapons

• Fraud and misleading financial offers

• Counterfeit goods

• Unsubstantiated health and wellness claims

• Certain medical products and treatments

• Legal services

• Individual job listings

• Individual housing listings

• Content featuring graphic violence or explicit sexual material

There are, however, some nuances. For example, underwear and swimwear may be permitted in a conventional retail or fashion context. A casino resort may potentially advertise a hotel stay if gambling is not what is being promoted.

🚫 Conversations where ads are not shown

Ad placement is just as important.

OpenAI states that ads are not placed alongside conversations involving topics such as:

• Personal health and mental health

• Suicide and self-harm

• Politics

• Children and child safety

• Sexual or graphic content

• Weapons

• Terrorism

• Hate and harassment

• Cyber abuse

• Fraud

• Illegal activities

• Privacy

• Dangerous activities

• Sensitive or emotionally vulnerable situations

This means that an otherwise permissible ad may be excluded from a particular conversation because the context is sensitive.

From an advertiser's perspective

One important detail is that OpenAI evaluates the entire ad experience – not only the ad creative itself. The landing page is also part of the review.

An ad that appears compliant in isolation may therefore be rejected if its landing page leads to content in a prohibited category.