OpenAI launches new visual ad format in ChatGPT as weekly users reach 1.2 billion
OpenAI is expanding advertising in ChatGPT with a new visual ad format, additional conversion and attribution integrations, incrementality testing and third-party brand-suitability pilots as the service reaches 1.2 billion weekly users.
The update was announced on October 5, 2026. The new format will initially be tested during image generation in ChatGPT, allowing advertisers to use visual creative alongside relevant generative workflows.
OpenAI says advertisements will remain clearly labeled, separate from generated images and unable to influence ChatGPT's answers. Testing begins later in October in the United States with an initial group of advertisers.
··········
THE CHATGPT ADVERTISING EXPANSION AT A GLANCE
........
Specification | Detail |
Announcement date | October 5, 2026 |
Platform | ChatGPT |
Weekly reach | 1.2 billion people |
New format | Visual advertising |
Initial placement | Image-generation workflows |
Initial market | United States |
Testing begins | Later in October 2026 |
Initial access | Selected advertisers |
Influence on answers | None, according to OpenAI |
Separation from output | Ads remain separate from generated images |
Measurement | Conversion, attribution and incrementality |
Brand-suitability pilots | DoubleVerify and Integral Ad Science |
........
The announcement adds several components required for ChatGPT Ads to operate as a measurable advertising channel rather than only a sponsored-placement experiment.
··········
VISUAL ADS WILL FIRST APPEAR DURING IMAGE GENERATION
OpenAI is starting with image generation, where the commercial format can show products, visual inspiration or experiences related to what the user is creating.
The placement is designed to remain distinct from the generated output. An advertiser can appear alongside the workflow, but its creative is not inserted into the image produced by ChatGPT.
That separation is especially important in generative interfaces because the assistant creates the primary content itself. OpenAI says paid placements will remain identifiable as advertising and will not alter the response generated by ChatGPT.
The initial test therefore introduces commercial imagery without turning the model's own output into sponsored content.
··········
1.2 BILLION WEEKLY USERS CREATE A LARGE NEW ADVERTISING MARKET
OpenAI says ChatGPT now reaches 1.2 billion people each week.
That audience gives the company a monetization opportunity substantially larger than the population willing to purchase a paid subscription.
The interaction model is also different from traditional digital advertising. Users can spend multiple turns defining requirements, comparing alternatives, researching products and refining a decision.
For advertisers, the relevant environment can therefore be an ongoing task or decision process rather than a single search query or a static webpage.
The commercial opportunity depends on maintaining a strict boundary between the advertising layer and the assistant itself. OpenAI's current policy is that ads are separate, labeled and do not influence answers.
··········
OPENAI IS CONNECTING CHATGPT ADS TO EXISTING MARKETING INFRASTRUCTURE
The October update also expands the systems advertisers can use to measure campaign performance.
OpenAI is adding integrations with Hightouch, Tealium and LiveRamp, enabling advertisers to send conversion information from their existing data infrastructure into ChatGPT Ads.
It is also working with a broader group of attribution and measurement providers.
........
Function | Partners named by OpenAI |
Conversion-data integrations | Hightouch, Tealium, LiveRamp |
Attribution and reporting | AppsFlyer, Triple Whale, Adjust, DV Rockerbox, Northbeam |
Additional attribution | Branch, Singular, Kochava, Airbridge, Tenjin |
Full-funnel measurement | Fospha, Measured, INCRMNTAL |
Incrementality testing | Haus, Measured, WorkMagic |
Brand suitability | DoubleVerify, Integral Ad Science |
........
This allows advertisers to compare ChatGPT campaigns with paid search, social advertising and mobile acquisition using measurement systems already present in their marketing stack.
··········
OPENAI IS TESTING WHETHER ADS CREATE INCREMENTAL SALES
Attribution identifies interactions associated with a conversion. It does not necessarily establish that the advertisement caused the conversion.
OpenAI is therefore working with Haus, Measured and WorkMagic on geo-based incrementality experiments intended to estimate how much additional demand ChatGPT Ads generate.
This distinction is important for large advertisers deciding whether to reallocate budget from established channels. A campaign can receive attribution for customers who would have purchased anyway, while incrementality attempts to measure the additional purchases created by the advertising exposure.
OpenAI describes this work as early-stage, so the results should not yet be treated as a general benchmark for ChatGPT advertising performance.
··········
EARLY ADVERTISER RESULTS SHOW LOWER CPA AND NEW-CUSTOMER ACQUISITION IN SOME CAMPAIGNS
OpenAI disclosed several early results reported through its measurement partners.
DV Rockerbox reported that WeightWatchers achieved a 15.3% lower attributed cost per acquisition through ChatGPT Ads than its blended paid-search benchmark.
WorkMagic reported statistically significant lift for wellness company Dose, with 67% of incremental purchases coming from new customers.
Triple Whale reported that 93% of Portland Leather visitors arriving through ChatGPT Ads were new visitors.
These are individual advertiser results rather than broad platform benchmarks. They indicate that OpenAI is already evaluating ChatGPT Ads against performance metrics normally used for mature acquisition channels, including CPA, incremental conversion and new-customer share.
··········
BRAND SUITABILITY HAS TO BE EVALUATED AT THE CONVERSATION LEVEL
Advertising inside ChatGPT introduces a placement problem that differs from conventional webpage advertising.
A webpage can often be classified before an advertisement is displayed. A conversation changes continuously, and its subject or emotional context can shift from one turn to another.
OpenAI says its placement systems are designed to avoid advertising in emotionally vulnerable, sensitive or otherwise unsuitable conversations.
The company is also introducing Negative Phrases for qualifying advertisers, allowing brands to define additional contexts in which their advertisements should not appear.
This creates two layers of control: OpenAI's platform-wide suitability rules and advertiser-specific exclusions.
··········
DOUBLEVERIFY AND IAS WILL TEST SAFETY WITHOUT ACCESS TO PRIVATE CHATS
OpenAI is also working with DoubleVerify and Integral Ad Science on independent brand-suitability pilots.
The testing model has an important privacy constraint: OpenAI says the companies will evaluate placement safeguards in a controlled environment without access to users' private conversations.
For advertisers, third-party verification can provide additional evidence that suitability controls are working. For users, the structure is intended to prevent external advertising-verification companies from inspecting the underlying private chat content.
This will be an important part of scaling conversational advertising because independent verification is standard practice for large advertising buyers, while ChatGPT conversations contain substantially more sensitive contextual information than conventional ad inventory.
··········
OPENAI IS BUILDING THE COMPONENTS OF A FULL ADVERTISING STACK
The October 5 update combines several layers that large advertisers generally expect from a mature platform:
visual creative → placement controls → conversion data → attribution → incrementality → brand suitability → third-party verification.
The new visual format expands what advertisers can show.
Hightouch, Tealium and LiveRamp connect conversion data.
Attribution providers measure customer journeys.
Haus, Measured and WorkMagic test incremental impact.
DoubleVerify and IAS address independent suitability verification.
The result is a more complete advertising infrastructure around ChatGPT, rather than a standalone sponsored-content format.
··········
THE MAIN CONSTRAINT IS PRESERVING THE BOUNDARY BETWEEN ADS AND THE ASSISTANT
At 1.2 billion weekly users, OpenAI already has sufficient scale for advertising to become a substantial business.
The more difficult problem is maintaining user trust while monetizing interactions that can involve research, purchasing decisions and personal context.
OpenAI's current design establishes three explicit boundaries: ads are labeled, ads remain separate from generated content and advertisers do not influence ChatGPT's answers.
The new visual format, conversion integrations, incrementality testing and brand-suitability partnerships show that OpenAI is now building the technical and measurement infrastructure required to scale advertising around those constraints.
The success of ChatGPT Ads will therefore depend not only on campaign performance, but on whether OpenAI can expand commercial inventory without making paid influence indistinguishable from the assistant's own output.
··········
FOLLOW US FOR MORE.
DATA STUDIOS
datastudios.org
