A visual ad format for ChatGPT image generation
OpenAI published Building advertising for the way people use AI on 5 October 2026. It introduces a new visual ad format in ChatGPT.
Testing will begin later this month in the US with an initial group of advertisers. Ads will appear during image generation, OpenAI says, clearly labeled and separate from the image being created.
OpenAI repeats that advertising does not influence the answers ChatGPT provides. Businesses can sign up at ads.openai.com.
OpenAI Ads and its measurement partners
OpenAI owns the format, the placement guardrails and the Ads product. The post cites weekly reach of 1.2 billion people on ChatGPT.
Measurement partners named for conversions and attribution include Hightouch, Tealium, LiveRamp, AppsFlyer, Triple Whale, Adjust, DV Rockerbox, Northbeam, Branch, Singular, Kochava, Airbridge and Tenjin. Incrementality partners named include Haus, Measured and WorkMagic.
Brand suitability pilots are listed with DoubleVerify and Integral Ad Science.
Images, measurement and suitability tools
The visual format is meant to show product inspiration, product usage or experiences. OpenAI says people can discover a brand and decide what to buy through those images.
Negative Phrases are now available to qualifying advertisers for brand-specific placement limits. Automated review, human oversight and monitoring are described as part of policy enforcement.
For earlier ChatGPT ads boundaries, see Where ChatGPT ads stop and advertiser messages begin.
Partner stats, US test only
Performance figures in the post are attributed to partners. DV Rockerbox is cited for a WeightWatchers cost per acquisition 15.3% below a blended paid-search benchmark. WorkMagic is cited for Dose, with 67% of incremental purchases from net-new customers. Triple Whale is cited for Portland Leather, with 93% of ChatGPT Ads visitors new.
Those are partner claims. The Frontier found no OpenAI methods paper for them in the post.
The visual format test is US only at launch. No price card appears in the announcement.
Advertisers and brand safety teams
- US advertisers already on ChatGPT Ads can ask for the visual format test cohort.
- Brand safety teams can review the DoubleVerify and IAS pilot design before widening spend.
- Publishers watching AI ad inventory can track whether image-generation placements stay separate from answer text, as OpenAI states.
