OpenAI Will Test Ads Inside ChatGPT Image Generation — and Opens Measurement to Outside Firms
In an Oct. 5 post, OpenAI said it will test a new visual ad format in ChatGPT during image generation with an initial group of U.S. advertisers later this month. Ads are labeled and kept separate from the generated image; OpenAI says they do not influence answers. The same update expands conversion, attribution and geo-incrementality partners, and starts brand-suitability pilots with DoubleVerify and Integral Ad Science in a controlled test environment without access to private conversations. Partner case studies — not audits — include WeightWatchers’ attributed CPA 15.3% below its paid-search blend per DV Rockerbox.
Ads are moving from the answer feed into the creative surface — and OpenAI is finally wiring the third-party measurement and brand-safety plumbing big brand budgets require.
In an October 5, 2026 post, OpenAI said it will test a new visual ad format in ChatGPT, starting during image generation, with an initial group of U.S. advertisers later this month. Ads will be clearly labeled and remain separate from the image being created. As with other ChatGPT ad formats, OpenAI says advertising does not influence the answers ChatGPT provides. The company repeats that ChatGPT reaches 1.2 billion people a week — its own figure.
What is shipping
The visual format is pitched as product inspiration, product-in-use, and experience imagery — a carousel-style placement next to generation, not inside the generated pixels. Placement still runs through OpenAI’s conversation guardrails that try to keep ads out of emotionally vulnerable or otherwise unsuitable contexts. Qualifying advertisers get Negative Phrases for brand-specific placement exclusions.
On measurement, OpenAI is opening the stack rather than asking brands to trust a closed dashboard:
- Conversion data integrations with Hightouch, Tealium and LiveRamp
- Attribution partners including AppsFlyer, Adjust, Branch, DV Rockerbox, Northbeam, Triple Whale, Singular, Kochava, Airbridge and Tenjin
- Early geo-based incrementality tests with Haus, Measured and WorkMagic
- Brand-suitability evaluation pilots with DoubleVerify and Integral Ad Science, run in a controlled test environment without access to private conversations
Claims vs checks
OpenAI cites partner-reported case studies, not independent audits. Per DV Rockerbox, WeightWatchers’ attributed cost per acquisition on ChatGPT Ads was 15.3% lower than its blended paid-search benchmark. WorkMagic reported statistically significant lift for wellness brand Dose, with 67% of incremental purchases from net-new customers. Triple Whale said 93% of Portland Leather visitors from ChatGPT Ads were new. Each figure is that partner’s attributed case study.
TNW notes ChatGPT ads have already run for Free and Go users in 31 European countries since late August — the Europe rollout Times of AI covered then — and that the EU’s Aug. 31 VLOSE designation requires a public ad repository from January. The earlier $1B annualized run-rate piece remains the desk’s cite for that OpenAI figure. Today’s visual-format post is not a new ARR print.
Limits
- Visual ads are a U.S. test with an initial advertiser group later in October — not general availability.
- Performance numbers are partner case studies quoted by OpenAI, not audited averages.
- Brand-safety pilots evaluate OpenAI’s guardrails in a sandbox, not live traffic with private chats.
- 1.2 billion weekly reach is OpenAI’s claim.