OpenAI has started placing advertisements directly inside ChatGPT’s responses, and the company has confirmed that the subject of a person’s conversation helps determine which ad shows up. The change ends nearly three years of ChatGPT running without any advertising, and it applies only to specific parts of the user base while OpenAI studies how people react to the new format. The company has published a detailed account of how the test works, including who sees ads, how they are chosen, and what advertisers can and cannot access.
The rollout began quietly in the United States in February 2026 and has since expanded internationally in stages. OpenAI has treated the effort as a staged pilot rather than a finished product, issuing periodic public updates as it adds countries, adjusts safeguards, and refines the system based on feedback from the people using it.
Free and Go Subscribers See Ads; Paid Tiers Do Not
The test applies to logged-in adult users on ChatGPT’s Free tier and its lower-cost Go subscription. According to OpenAI, subscribers on the Plus, Pro, Business, Enterprise and Education tiers do not see ads at all under the current test. People who would rather avoid ads entirely can upgrade to one of those paid plans, or, while staying on the Free tier, opt out of ads in exchange for a smaller number of daily free messages.
Conversation Topics and Past Chats Shape Which Ad Appears
During the test, OpenAI selects which advertisement to display by matching submissions from advertisers against the subject of the ongoing conversation, a person’s past chats, and how that person has previously interacted with ads. The company offers its own example: someone researching recipes for a gathering might be shown an ad for a meal-kit delivery service. When more than one advertiser’s product could plausibly fit a conversation, the system is designed to pick whichever is judged most relevant and surface that one first.
Advertisers Receive Click and View Counts, Not the Conversation
Even though the content of a conversation helps choose the ad, OpenAI says advertisers are not given access to chats, chat history, memories, or other personal details. What advertisers receive instead is aggregate performance information, such as how many times an ad was viewed or clicked across all the people who saw it. OpenAI has also said the presence of ads does not change the substance of ChatGPT’s answers: the company states that answers remain independently generated, and that any ad appearing alongside one is visually separated from the response and clearly labeled as sponsored.
Minors and Sensitive Subjects Are Walled Off From the Test
Accounts that OpenAI identifies, or predicts, belong to someone under 18 are excluded from seeing ads entirely during the pilot. The company has also kept ads from appearing near topics it treats as sensitive or regulated, including health, mental health and politics. OpenAI describes these exclusions as safeguards that will continue to develop as the pilot matures, language that suggests the current boundaries are a starting point rather than a fixed, permanent policy.
Dismissing Ads and Deleting the Data Behind Them
People inside the test group have several direct controls over the ads they see. They can dismiss an individual ad, submit feedback about it, view an explanation of why that specific ad was selected for them, and delete all of the ad-related data tied to their account in a single action. A separate settings panel lets someone turn ad personalization off altogether, including whether the system is allowed to draw on past chats and saved memories when choosing what to display next.
A U.S. Pilot That Has Already Reached Five More Countries
OpenAI frames the shift as a funding question rather than a purely commercial one. The company says ChatGPT is now used by hundreds of millions of people for learning, work and everyday decisions, and that keeping the Free and Go tiers fast and reliable requires substantial, ongoing infrastructure investment. Advertising, in OpenAI’s telling, is meant to help fund that infrastructure so more people can use higher-quality free or low-cost versions of the product, while the company keeps expanding the underlying models it offers over time. OpenAI has also said it expects advertising to make up less than half of its total revenue over the long term, positioning ads as one funding stream among several rather than the company’s primary business.
OpenAI began testing ads with Free and Go users in the United States on February 9, 2026. By late March, the company said it would expand into Canada, Australia and New Zealand. A May update announced plans to extend the pilot further, to the United Kingdom, Mexico, Brazil, Japan and South Korea, and by OpenAI’s own account, the ad test had formally launched in all five of those additional markets by August 11, 2026, with further expansion expected before the end of the year. Businesses interested in placing ads inside ChatGPT can register their interest through OpenAI’s advertiser sign-up page, which the company points to for updates as the program adds new formats and buying options over time.
This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.
More from Morning Overview
- Consumer Reports names the 2026 models it expects to break down the most
- Supplements now rank as the fifth-leading cause of death from liver disease.
- The NSA is again telling phone owners to switch off one location setting
- Four U.S. startups fired up their first small nuclear reactors, aiming to power AI data centers on-site
