OpenAI on October 5, 2026 introduced a new visual ad format in ChatGPT, expanded the measurement tools and partner ecosystem behind ChatGPT Ads, and said it is developing brand suitability evaluation pilots with DoubleVerify and Integral Ad Science, saying the platform reaches 1.2 billion people each week.
Image Generation Test and Ad Labeling
According to OpenAI, the new format uses images showing product inspiration, product usage, or the experiences a product makes possible, so people can discover new brands, see what sets an offering apart, and make purchase decisions. The company said the format creates new opportunities for advertisers to tell their stories through images.
OpenAI will initially test the format during image generation in ChatGPT, and the company said testing will begin later in October 2026 in the US with an initial group of advertisers. OpenAI said the ads will carry clear labels and stay separate from the image a user is generating, and that, as with its other ad formats, ads have no influence over ChatGPT’s answers.
The company described ChatGPT as the world’s largest AI-native consumer platform. OpenAI said advertising allows it to serve a much larger market than subscriptions alone, and gives businesses a new way to reach people while they explore ideas, make decisions, and find products and services tied to their goals.
Measurement Partners and Early Results
OpenAI said its measurement approach for ChatGPT Ads supports both in-house and partner-led solutions, covering signals from its pixel and Conversions API, attribution, incrementality, and brand measurement, and it named a first set of third-party partners across those offerings. A companion post on OpenAI’s Ads Blog published the same day details the mechanics.
Integrations with Hightouch, Tealium, and LiveRamp let advertisers send conversion data from their existing systems to ChatGPT Ads. The pixel and Conversions API measure actions such as purchases, leads, installs, and sign-ups, and those signals are used to optimize campaigns. A feature called Event Quality Score lets advertisers gauge the quality of their conversion signals and spot ways to improve them.
Supported attribution partners across the web and app space include AppsFlyer, Triple Whale, Adjust, DV Rockerbox, Northbeam, Branch, Singular, Kochava, Airbridge, and Tenjin, enabling Conversions API, reporting, and click attribution solutions. Fospha, Measured, and INCRMNTAL serve as full-funnel and advanced measurement partners. OpenAI is also working with Haus, Measured, and WorkMagic on geo-based incrementality experiments aimed at measuring the causal impact of ChatGPT Ads, work it characterized as early-stage, and is separately in early-stage testing with Kantar and Cint on measuring how ChatGPT Ads affect brand resonance metrics.
Advertisers can choose a click-through attribution window matched to their buying cycle and add a one-day view-through window for conversions that occur after an impression without a click. According to OpenAI, those settings affect only attribution in reports and do not alter how campaigns optimize. The company said its advanced conversion optimization draws on eligible click-through and view-through conversions for an advertiser’s chosen goal, so purchases that follow either path can shape future ad delivery.
OpenAI said that in an early analysis of global campaigns optimized for deeper-funnel outcomes, 52.7% of eligible one-day view-through conversions took place within an hour of the matched ad impression, and that it already sees conversions following ad impressions even when there is no qualifying click.
OpenAI also cited early partner-reported performance. According to 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 coming from net-new customers, and a WorkMagic geo-lift study estimated 2.3x more incremental orders for Dose than last-click attribution captured. According to Triple Whale, 93% of Portland Leather’s visitors from ChatGPT Ads were new.
Brand Suitability Pilots and Placement Guardrails
OpenAI said its advertising policies and guardrails exist to protect users, preserve trust, and give brands confidence in where their ads appear. The company said its placement guardrails assess whether a conversation is appropriate for advertising, with the aim of keeping ads out of contexts that are emotionally vulnerable, sensitive, or otherwise unsuitable. Automated review, human oversight, and ongoing monitoring are used to identify and address policy violations and improve safeguards over time. Negative Phrases, which address narrow placement requirements tied to a brand’s own policies, are now available to qualifying advertisers.
Under the pilots, DoubleVerify and Integral Ad Science will evaluate how OpenAI applies its brand safety standards inside a controlled testing environment, without access to private user conversations. OpenAI said the pilots are meant to give advertisers more visibility into how the safeguards operate while keeping conversations private.
Integral Ad Science CEO Lidiane Jones said her firm’s independent evaluation will assess OpenAI’s safeguards across a wide and evolving set of scenarios and give advertisers more transparency into brand safety and suitability without exposing real user conversations. DoubleVerify CEO Mark Zagorski said advertising in ChatGPT differs from web and video placements because the conversational context, rather than a page or video, determines where an ad appears, and he said DoubleVerify intends to bring independent measurement to that environment while respecting user privacy.
OpenAI said the updates are consistent with its ads principles: ChatGPT’s answers stay independent, conversations stay private, and users remain in control of their experience. The company said it will continue to evolve the ad platform with new formats and capabilities, and businesses can sign up for ads at ads.openai.com.
Credit: Source link
