OpenAI Will Test Visual Product Ads in ChatGPT During Image Generation

OpenAI said on October 5, 2026 that it will begin testing visual ads with product images inside ChatGPT later this month, starting in the US. The first placement is the screen users see while ChatGPT generates an image, where labeled product ads will sit below the image being created.

OpenAI has sold ads in ChatGPT since February. According to Digiday, that format combines an advertiser name, logo, headline, description, landing page and image; the new one puts product pictures at the center. TechCrunch notes that the move follows the launch of Meta's Muse, a free agentic AI app that competes with ChatGPT for the same users. We track both in products and tools.

What exactly did OpenAI announce?

OpenAI announced a new visual ad format for ChatGPT, a set of new measurement partners and two brand safety pilots, in a blog post titled "Building advertising for the way people use AI" on October 5, 2026. The visual ads show product images with a "Learn more" button.

According to OpenAI, the ads will be clearly labeled and kept separate from the image ChatGPT is creating. The Decoder describes the format as a carousel or a row of product images below the generated picture, meant to show products in use. The Verge reports that OpenAI says the ads will "initially" appear during image generation and that other formats will follow later.

Engadget quotes OpenAI's post: "As with our other ad formats, advertising does not influence the answers ChatGPT provides." That is a company claim; no outside audit of it has been published.

Who will see the ads, and when?

US ChatGPT users will be first, with the test starting later in October 2026 and limited to an initial group of advertisers, OpenAI said. The company did not say how many advertisers are in that group or when the format will reach other countries.

OpenAI also says it decides when a conversation is appropriate for ads. The Verge reports that the goal is to keep ads out of "emotionally vulnerable, sensitive, or otherwise unsuitable contexts", and Engadget gives personal health and mental wellbeing as examples of topics where ads will not show.

For scale, OpenAI says ChatGPT reaches 1.2 billion people every week. TechCrunch notes that OpenAI introduced ads earlier in 2026 to support its free and low cost subscription tiers and expanded them to India in August.

How will advertisers measure results?

OpenAI expanded its measurement tools so advertisers can send conversion data and use attribution tools they already work with, the company said on October 5, 2026. Data can flow in through Hightouch, Tealium and LiveRamp.

The attribution partners OpenAI lists are AppsFlyer, Triple Whale, Adjust, DV Rockerbox, Northbeam, Branch, Singular, Kochava, Airbridge and Tenjin. Fospha, Measured and INCRMNTAL are named for full funnel measurement. Haus, Measured and WorkMagic will help advertisers run geo based experiments, a method that compares regions with and without ads to estimate how many sales the ads actually caused.

Digiday reports that measurement has been one of the main complaints from advertisers since February, along with too few formats and too little inventory. OpenAI's post also cites results from advertisers it worked with, including WeightWatchers, which OpenAI says saw a 15.3% lower cost per acquisition than blended paid search. Those figures come from OpenAI and the advertisers, not from an independent study.

What changes for brand safety?

DoubleVerify and Integral Ad Science are piloting brand suitability evaluations for ChatGPT ads, according to OpenAI. Brand suitability tools tell an advertiser whether the content next to an ad fits its brand.

The Decoder reports that these pilots evaluate the ad environment without access to real user conversations. OpenAI also introduced "negative phrases" for qualifying advertisers, which keep ads away from conversations an advertiser wants to avoid, similar to negative keywords in search advertising.

How big is ChatGPT's ad business?

OpenAI's ad business is still young. The Decoder reports that OpenAI has been running ChatGPT ads since February, now in more than 40 countries, and that the company has reached an annualized revenue run rate of $1 billion from ads alone. OpenAI's October 5 post does not repeat those figures.

Digiday points to a structural limit: ChatGPT has fewer ad placements than search engines, so supply may lag behind what advertisers want to spend. The visual format adds a new placement without changing what ChatGPT says in its answers.

What we don't know yet

• How many advertisers are in the initial test group, and which brands they are.
• When visual ads will reach countries outside the US, including the EU.
• Which ChatGPT plans will see the visual ads, beyond the free and low cost tiers that already carry ads.
• How OpenAI will price the new format compared with its text ads.
• Whether independent auditors will verify that ads do not shape ChatGPT's answers.