ChatGPT Is Testing Ads: What's Changing for Free Accounts
OpenAI is testing ads in ChatGPT for Free and Go accounts. Here's how it works, how targeting is handled, what privacy guarantees are in place, and what it means for users.
Summary
Since February 9, 2026, OpenAI has been testing ads in ChatGPT to help fund free access while aiming to preserve user trust. The pilot launched in the US and, as of August 2026, has expanded to the United Kingdom, Mexico, Brazil, Japan, and South Korea. Only Free and Go accounts (adults) see ads, clearly labeled "sponsored" and displayed as cards beneath the chatbot's responses. Plus, Pro, Business, Enterprise, and Education subscribers remain ad-free. OpenAI highlights three main safeguards: advertisers never get access to conversations, chat history, or personal data; ChatGPT's answers stay independent and are never influenced by advertising; and users get tools to personalize, hide, or clear their ad history.
Why OpenAI is bringing ads to ChatGPT
With hundreds of millions of active users, ChatGPT is still used for free by the vast majority of them. To keep offering free access while covering the massive infrastructure and compute costs behind AI models, OpenAI is following the same path already taken by Google or Meta: advertising. The company frames this pilot as a way to "help sustain free access" to ChatGPT without compromising the experience or the neutrality of the model's answers.
How ads work in ChatGPT
In practice, ads appear as product cards displayed below ChatGPT's response, visually separated from the AI-generated content and labeled "sponsored." Multiple advertisers can appear at the same time within a single conversation. During this test phase, ads do not appear in the ChatGPT Atlas browser.
Ad targeting relies on:
the context of the current conversation,
general location and language,
and, if enabled, past chat history and "memories" saved by ChatGPT.
OpenAI states that this targeting information "stays within ChatGPT" and is not shared with advertisers, who only receive aggregated metrics (impressions, clicks) — never the content of conversations. In the European Economic Area and Switzerland, ad personalization is not initially enabled — a distinction that directly matters for users in those regions.
Who sees ads, and who doesn't
Plan | Ads |
|---|---|
Free | Yes |
Go | Yes |
Plus | No |
Pro | No |
Business | No |
Enterprise | No |
Education | No |
Users under 18 receive no ads at all, based on account-level age data and predictive modeling. OpenAI also excludes ads near sensitive topics such as personal health, mental health, or politics — political advertising is currently prohibited outright.
The guarantees OpenAI is emphasizing
Aware of how sensitive ad monetization is for a conversational assistant, OpenAI highlights several commitments:
Conversation privacy — advertisers have no access to conversations, chat history, memories, or personal information. Data is never sold.
Independence of responses — ChatGPT's answers remain, according to OpenAI, "independent and unbiased": the presence of an advertiser never influences the content or wording of a response.
User control — anyone can turn off ad personalization, hide or report an ad, view their ad history, clear it (retained for 30 days), or switch to an "ads-free" mode with reduced message limits. The most direct option remains upgrading to a paid plan.
The takeaway
This pilot marks a turning point in ChatGPT's business model, aligning it with the major ad-funded platforms. For free-tier users, the change is tangible: sponsored cards will start showing up under certain responses. As the rollout expands to more countries, the real test will be whether the promised guarantees around response independence and data protection hold up over time — and whether the user experience stays true to what made ChatGPT successful in the first place.
Sources: Testing ads in ChatGPT — OpenAI, Ads in ChatGPT — OpenAI Help Center
