chatgpttutorialshow-toai

Spot the New ChatGPT Image Ads on Free and Go

Learn how ChatGPT image ads work on Free and Go, which plans stay ad-free, and how to dismiss, review, or limit ads.

October 9, 2026 Yum News

OpenAI is testing a new visual ad format next to image generation in ChatGPT. The test starts later in October 2026 in the United States, and it is limited to logged-in adults on the Free and Go plans.

If you use ChatGPT to make pictures, a labeled ad may appear beside the image you asked for. It is not part of that image. This guide explains who sees the ads, how OpenAI says they are separated from answers, and what controls you already have.

What OpenAI announced on October 5

On October 5, 2026, OpenAI said it is introducing a visual ad format that shows product inspiration, product use, or related experiences. The goal is to help people discover a brand and decide what to buy, while giving advertisers a way to tell a story with images.

The first test runs during image generation in ChatGPT for Free and Go users. OpenAI says ads will be clearly labeled and will stay separate from the image being created. Advertising does not influence the answers ChatGPT provides. Testing begins later this month in the US with an initial group of advertisers.

OpenAI also said ChatGPT reaches 1.2 billion people each week. That figure is the company’s own reach claim, not an independent audit.

Person reviewing a laptop and phone at a desk

Who sees ads, and who does not

OpenAI started testing ads in the US on February 9, 2026, for logged-in adult users on Free and Go. Plus, Pro, Business, Enterprise, and Education plans do not get ads.

The October visual format follows the same split. If you pay for Plus or Pro, or you use a work or school plan, this image-ad test should not appear in your account.

Accounts where the user says they are under 18, or where OpenAI predicts they are under 18, are not shown ads. OpenAI has also said ads are not eligible to appear near sensitive or regulated topics such as health, mental health, or politics. Placement guardrails are meant to keep ads out of emotionally vulnerable or otherwise unsuitable conversations.

If you are on Free and would rather not see ads, OpenAI’s February post said you can opt out in exchange for fewer daily free messages, or upgrade to Plus or Pro.

How an image ad is supposed to look

OpenAI’s product note is specific about placement. The ad sits with the image-generation experience, but it is not mixed into the picture ChatGPT makes for you.

Use this checklist when you generate an image on Free or Go in the US:

Finish the image prompt as usual. The picture you requested is the organic result.
Look for a separate card or image unit near that result, not inside the generated file.
Confirm a clear sponsored or ad label. OpenAI says ads are always labeled and visually separated from the organic answer.
Treat the labeled unit as optional. It is a product or service suggestion, not part of the prompt result.
Do not assume the brand in the ad shaped the answer. OpenAI says answers stay independent of advertising.

If a brand mark appears inside the picture you asked ChatGPT to create, that is a different issue from this ad format. The October test is about a labeled unit kept apart from the generated image.

Why a particular ad can appear

During the earlier ads test, OpenAI said it matches advertiser submissions to the topic of the conversation, past chats, and past ad interactions. A recipe chat might surface meal-kit or grocery ads. If several advertisers qualify, OpenAI shows the one it judges most relevant first.

Advertisers do not receive your chats, chat history, memories, or personal details. They get aggregate performance data such as views or clicks.

The October update also expands measurement for businesses. OpenAI named conversion and attribution partners including Hightouch, Tealium, LiveRamp, AppsFlyer, Triple Whale, Adjust, and others, plus early incrementality work with Haus, Measured, and WorkMagic. Those tools are for advertisers. They do not change the rule that chats stay private from advertisers.

Partner-reported examples in the same post include a lower attributed cost per acquisition for WeightWatchers versus its paid-search benchmark (DV Rockerbox), lift for Dose (WorkMagic), and a high share of new visitors for Portland Leather (Triple Whale). Those numbers come from the partners, not from an OpenAI benchmark table.

Analytics charts on a laptop screen

Controls you can use today

OpenAI’s ads principles say people keep control of the experience. In the February launch note, the company listed these actions:

Dismiss an ad you do not want.
Send feedback on an ad.
Open the explanation for why that ad was shown.
Delete your ad data in one tap.
Manage ad personalization at any time.

Exact menu names can move as the app updates. Start in ChatGPT settings and look for ads, personalization, or data controls. If you already use the privacy center, review ad data there alongside chat history settings. Our walkthrough of the ChatGPT privacy center covers the broader data controls that sit next to these ad options.

Negative Phrases, a placement control for qualifying advertisers, is not a user setting. It lets certain brands block narrow contexts that conflict with their own policies.

What the US image test does not include yet

The visual format is a limited US test that starts later in October 2026. It is not a global rollout on day one. OpenAI has not published a public list of the first advertisers.

Other ad work is separate. In September 2026, OpenAI said it was testing Sponsored Agents with select US advertisers, so a click can open a conversation with a business-sponsored agent. That is not the same as the image-generation unit announced on October 5.

Brand-suitability pilots with DoubleVerify and Integral Ad Science are also for advertisers. OpenAI says those partners assess safeguards in a controlled test environment and do not access private user conversations.

The video above is OpenAI’s Ads Manager walkthrough for businesses creating a campaign at ads.openai.com. It is useful if you want to see how an advertiser sets a destination URL and context hints. It does not replace the in-app controls for people who only want to manage the ads they see.

Practical tips before the test widens

Check your plan first. Free and Go are in scope for ads. Plus, Pro, Business, Enterprise, and Education are not.

Generate a test image after the US rollout window and look only for a labeled unit beside the result. If nothing labeled appears, you may be outside the initial advertiser group, outside the US, or on a plan that does not get ads.

Keep sensitive topics in mind. OpenAI says ads should not run next to health, mental health, or politics. If you see a sponsored unit in one of those chats, use the feedback control and dismiss it.

If you rely on ChatGPT for client work, use a plan that OpenAI says does not include ads. That avoids a labeled product card sitting next to a generated asset you might export.

Bottom line

The October 2026 image ad is a labeled unit shown during image generation for Free and Go users in a US test. It is meant to stay separate from both the answer and the picture you requested. Paid Plus and Pro plans, plus Business, Enterprise, and Education, remain outside the ads test described by OpenAI.

You can dismiss an ad, ask why it appeared, delete ad data, and manage personalization. If you want fewer interruptions on Free, OpenAI’s earlier option is to opt out and accept fewer daily free messages.