OpenAI tests ChatGPT visual ads in late October

OpenAI's move puts advertising directly inside one of ChatGPT's most visual surfaces: the moment a user is generating an image. The company is pairing that placement with the plumbing advertisers expect before they spend — conversion data flowing in from their existing systems, reporting, and tracking from an ad click through to a result.

Two details in the arrangement are worth noting. The first is the separation between the ad and the generated image, with the ad clearly labeled; OpenAI says the ads do not influence ChatGPT's answers. The second is the control layer around where ads appear. OpenAI is offering an exclusion-phrase feature to advertisers that meet its conditions, and is developing a trial program with DoubleVerify and Integral Ad Science to evaluate conformity with brand placement standards. That evaluation is designed to happen in a controlled test environment, without access to users' private conversations.

The test's significance likely hinges on whether brands trust the separation between paid placement and model output, and on whether the conversion integrations prove the format actually drives purchases. Advertisers with strict placement criteria will be watching the exclusion-phrase tool and the DoubleVerify and Integral Ad Science trial closely.