OpenAI says it plans to test visual ads during image generation in ChatGPT, giving advertisers another place to reach people using the chatbot. The proposed format would show products in use or illustrate experiences that businesses offer, while keeping the advertisement separate from the image a user is creating.
The company has outlined an initial US test with a limited group of advertisers. It says the placements would carry clear labels and that advertising does not influence ChatGPT’s answers. Those are commitments about how the experience should work, rather than evidence of how reliably it will work in practice.
OpenAI presents advertising as a way to fund broader access to ChatGPT than subscriptions alone can support. The visual format would extend that business into the experience of making images.
How OpenAI describes the safeguards
OpenAI says its placement checks are designed to keep ads out of conversations involving emotional vulnerability or sensitive subjects. That makes the conversation’s suitability for advertising part of the decision about whether an ad should appear.
The practical question is how consistently those checks recognize situations where a commercial message would be inappropriate. A stated policy establishes the intended boundary; evaluating its performance is a separate task.
OpenAI says it is developing evaluation pilots with DoubleVerify and Integral Ad Science. The proposed arrangement would let those companies examine its brand safety standards in a controlled environment without access to private user conversations. A firm timeline for those evaluation pilots has not been publicly confirmed.
Europe adds advertising transparency requirements
OpenAI describes its European advertising expansion as covering 31 countries, with placements limited to Free and Go users. Its stated policy keeps Plus, Pro and Enterprise subscriptions free of ads. That broader advertising rollout is separate from the proposed US test of the visual format.
The European Commission identifies ChatGPT as a very large online search engine under the Digital Services Act. It announced that designation on August 31, 2026, and says the additional obligations must be met by January 2027. Its register lists 159.1 million monthly EU users for ChatGPT, above the 45 million threshold for designation.
For advertising, the law requires designated large search engines and platforms to maintain a searchable public repository. An ad’s information must remain accessible throughout its run and for a year after its last appearance.
The records must identify the advertiser and any separate payer. They must also disclose when the ad appeared, its reach and the main parameters used to select or exclude particular audiences. The repository must not contain the personal data of people who received the ads.
The Commission also names risks to users’ physical and mental well-being among the issues designated services must assess and mitigate. For ChatGPT, that puts the handling of sensitive conversations alongside a concrete transparency obligation: making its advertising activity open to public scrutiny.
