OpenAI is giving ChatGPT ads a visual layer. On October 5, the company said it will test a picture-led ad unit inside the app’s image-generation flow, starting later this month with a small group of US advertisers. It also expanded ChatGPT’s measurement vendor list to 18 companies and named DoubleVerify and Integral Ad Science for brand-safety evaluation pilots.
The placement: image generation becomes ad inventory
The new unit is built around imagery that shows product inspiration, product use or the experience a product enables. A mock-up in OpenAI’s announcement shows a grocery brand with a dinner-themed headline, a photograph and a Learn more button. The company says ads will be clearly labelled and kept separate from the image being created, and that advertising does not influence ChatGPT’s answers.
For now, the trial is limited to the United States. OpenAI gave no pricing, no list of participating advertisers and no indication whether the unit will appear in the self-serve Ads Manager.
Why the scale changes the brief
OpenAI says ChatGPT reaches 1.2 billion people each week, up from 900 million disclosed in February. The company frames ads as a way to serve a much larger market than subscriptions alone. For media buyers, that audience is the real reason to care: ChatGPT’s ad business already passed a $1 billion annualised run rate on August 31, from a pilot that began at a $60 CPM in February.
Measurement grows, but the proof is still early
The expanded vendor roster includes data connections through Hightouch, Tealium and LiveRamp, ten attribution partners, and incrementality work with Haus, Measured and WorkMagic. Three early partner-reported results were shared:
- WeightWatchers’ attributed cost per acquisition on ChatGPT Ads was 15.3% lower than its blended paid-search benchmark, according to DV Rockerbox.
- WorkMagic reported that 67% of incremental purchases for wellness brand Dose came from net-new customers.
- Triple Whale said 93% of Portland Leather’s visitors from ChatGPT Ads were new.
Treat all three with care: none comes with spend, dates, sample sizes or confidence intervals, and each reaches readers through a vendor-reported chain.
Brand suitability gets third-party pilots
OpenAI says placement guardrails aim to keep ads away from emotionally vulnerable or otherwise unsuitable conversations, and a new Negative Phrases control is available to qualifying advertisers. DoubleVerify and Integral Ad Science are developing evaluation pilots that will assess brand-safety standards without accessing private conversations. IAS chief Lidiane Jones said marketers need confidence that brand safety protections are durable, consistent and built to operate at scale.
What to do before you buy
Scale alone is not proof. When the trials open, ask for the method behind any lift figure, request spend and date ranges, and test the visual unit with a holdout rather than shifting budget on a supplier’s number alone. If you run a brand in a sensitive category, ask how Negative Phrases work and whether image prompts can trigger your ads.
Source: PPC Land
