Jakarta, ThedailyID — OpenAI is preparing to test a new visual advertising format on ChatGPT. The ads will appear when users generate images through the chatbot.
The company plans to begin testing the format in the United States in late October 2026. The initial test will involve selected advertisers.
The ads will appear beside the images requested by users. OpenAI says the ads will carry clear labels and remain separate from the generated content.
The new format can display product images, demonstrations, and experiences offered by a product or service. OpenAI says the format could help users discover brands and understand products before making a purchase.
OpenAI also says advertisements will not influence ChatGPT’s responses. The company maintains that its advertising system will remain separate from the chatbot’s answers.
“We are introducing a new visual advertising format in ChatGPT,” OpenAI said in a statement published on its official website on October 5, 2026.
Alongside the new format, OpenAI is expanding its measurement system for ChatGPT Ads. The company is working with Hightouch, Tealium, and LiveRamp.
These partnerships will allow advertisers to send conversion data from systems they already use to ChatGPT Ads. OpenAI says the system can help advertisers measure campaign performance.
The company also shared early testing results from its advertising system. WeightWatchers, for example, reportedly reduced customer acquisition costs by 15.3 percent.
Fitness brand Dose, meanwhile, reportedly generated 67 percent of its purchases from new customers during testing.
OpenAI is also strengthening its brand suitability systems. The company says its policies and systems aim to prevent ads from appearing in sensitive or inappropriate conversations.
The safeguards also cover conversations involving users in emotionally vulnerable situations. OpenAI says advertising should not take advantage of those circumstances.
The company is testing brand suitability evaluations with DoubleVerify and Integral Ad Science, or IAS. Independent partners can assess OpenAI’s brand safety standards in a controlled environment.
OpenAI says those evaluations do not give the partners access to users’ private conversations.
