OpenAI will test visual ads during ChatGPT image generation

OpenAI will test visual advertisements during ChatGPT image generation in the US later in October. Its announcement on 5 October 2026 describes a format for product inspiration and says ChatGPT now has 1.2 billion weekly users.

The initial test will involve a group of US advertisers. OpenAI says the ads will be clearly labelled, shown separately from the generated image and kept independent of ChatGPT’s answers. The audience figure is the company’s own reported count.

The sponsored placement sits beside the image

The new visual format is designed for image-generation experiences in which people seek inspiration for products or activities. OpenAI’s announcement keeps the sponsored placement and generated image separate, so a reader can identify the advertisement as its own element.

OpenAI says advertising cannot influence the answers ChatGPT provides. The company presents those boundaries as product commitments for the test.

Free and Go are the advertising-supported plans

OpenAI’s ChatGPT advertising guidance identifies Free and Go as plans that can carry ads. Plus, Pro, Business, Enterprise and Edu remain ad-free.

The guidance also describes user controls, including dismissing or reporting an ad and managing advertising personalisation. OpenAI says advertisers receive aggregate performance information rather than access to individual conversations. These rules provide the context for the forthcoming visual test.

Advertisers are getting more measurement options

OpenAI is expanding the systems advertisers can use to connect a ChatGPT ad exposure to a business result. The measurement programme includes conversion-data integrations and partners for attribution, full-funnel assessment and incrementality testing.

Attribution estimates which advertising interactions contributed to an action such as a purchase. Incrementality asks a different question: how much additional activity the advertising caused. OpenAI names Haus, Measured and WorkMagic for geographic experiments that can help compare advertising effects across areas.

The announcement also names DoubleVerify and Integral Ad Science for an initial brand-suitability pilot. OpenAI says that work uses controlled testing without giving those partners access to private user conversations.

The visual test extends an existing ad pilot

Advertising agency Dentsu announced its participation as an early OpenAI ad-pilot partner. It described the programme as an opportunity to test how brands appear in conversational experiences.

The October announcement adds a visual format and a larger measurement toolkit to that developing advertising business. The US test will show how a separately labelled product placement works alongside the act of generating an image.