OpenAI will begin testing visual advertisements alongside ChatGPT image generation in the United States later this month. Sponsored products or services may appear as a row or carousel while an image is being generated, expanding beyond the company name, logo and link used in earlier text ads.
The company says the placements will be labeled, remain separate from the generated image and will not influence ChatGPT’s answers. Plus, Pro and Enterprise subscribers will not see the ads. OpenAI plans to start with a limited group of advertisers and test additional formats later.
OpenAI also says it uses suitability controls intended to keep ads away from emotionally vulnerable, sensitive or otherwise inappropriate conversations. It is piloting independent brand-safety evaluation with DoubleVerify and Integral Ad Science, which the company says can assess the advertising environment without reading actual user conversations.
The format gives OpenAI another way to fund free and lower-cost access, but it also brings ordinary ad-platform concerns into a conversational product. Clear separation does not resolve questions about targeting, measurement or whether users understand how interaction data affects ad selection. The initial US test will show the presentation; fuller privacy and effectiveness judgments require the final product documentation and behavior.