ChatGPT Ads Launch in Europe: Who Will See Them and How to Go Ad-Free

During the week beginning August 24, OpenAI is set to expand ChatGPT Ads to 31 European markets. Ads may appear below answers for people on Free and Go, while Plus, Pro, Business, Enterprise, and Edu accounts are supposed to remain ad-free. There is an important distinction for users in the European Economic Area and Switzerland: personalized ads are not initially available, but ads may still be selected from the context of the current conversation.

This is more than another place to put a banner. ChatGPT is gaining an advertising system next to conversations in which people may already be comparing products, planning travel, or choosing software. This guide separates what OpenAI has confirmed from what remains untested, explains what advertisers are not supposed to see, and shows where to look for Ad Controls as the European rollout begins.

What OpenAI announced and when the rollout begins

In an announcement published on August 18, OpenAI said ChatGPT Ads would expand to 31 European markets “next week”—the week of August 24–30. It specifically mentioned Germany, France, Spain, Italy, Sweden, Norway, Denmark, the Netherlands, and Austria. This is the platform’s largest geographic expansion since the US pilot began in February 2026 and ads subsequently reached eight more markets.

“Next week” should not be read as a promise that every eligible account will display ads at 12:01 a.m. on Monday. OpenAI’s separate Ads Manager availability table describes self-service access for advertisers, not ad delivery to users; European countries are still marked “Coming Soon” there. The advertiser dashboard and consumer rollout therefore cannot be checked through the same status.

The upcoming advertiser list includes Greece, Poland, and Slovakia, among others. Ukraine is not in the current table. However, that page describes access to Ads Manager and can change as testing expands; it is not a permanent list of every country where a user may or may not see an ad in the future.

Who may see ads and which plans remain ad-free

According to OpenAI’s advertising FAQ, ads may appear for people on Free and Go. The company lists Plus, Pro, Business, Enterprise, and Edu as ad-free. OpenAI also says that it does not show ads to accounts it identifies as belonging to people under 18.

Plan or mode	Ads	What matters
Free	May appear	A separate ad-free option may be available with lower limits and fewer tools.
Go	May appear	Disabling personalization does not remove contextual ads.
Plus / Pro	No	OpenAI describes these paid subscriptions as ad-free.
Business / Enterprise / Edu	No	Organization and education accounts are not supposed to show ads.
Temporary Chat	No	Temporary chats do not show ads during the test.

Not every eligible account will necessarily show an ad at the start of the week. OpenAI describes the program as a test and a gradual expansion. If Ad Controls or sponsored units are absent on August 24, that does not mean the country has been excluded permanently. Updating the app and checking the account later is more useful than treating Monday as a universal deadline.

How an ad is separated from a ChatGPT answer

OpenAI says its advertising system is separate from the chat model and that advertisers cannot shape, rank, or alter ChatGPT’s answers. A sponsored unit may appear below a completed answer, where it should be clearly labeled and visually distinct from the model’s response. A paid placement also does not mean OpenAI endorses or recommends the advertiser, its product, or its claims.

The boundary matters, but people still need to read the interface carefully. A relevant offer shown after a laptop comparison can feel like a natural continuation of the conversation. Check the sponsorship label, destination domain, pricing conditions, and independent reviews just as you would in search or on social media. If the service itself is still unfamiliar, start with our practical introduction to ChatGPT.

What advertisers do not see, according to OpenAI

OpenAI’s Help Center states that advertisers do not receive your chats, chat history, memories, name, email address, precise location, IP address, or sensitive personal information. Advertisers receive aggregated information about campaign performance, such as total impressions and clicks. If a user chooses to message an advertiser through an ad, that advertiser sees only the messages sent directly to it.

This is OpenAI’s description of its policy, not an independent audit of every component in the advertising stack. “Not shared with an advertiser” also does not mean “not processed by the platform.” To decide which ad is relevant, ChatGPT’s advertising system still evaluates certain signals internally. That is why contextual selection, personalization, and advertiser reporting need to be treated as three different things.

Contextual and personalized advertising are not the same

Contextual advertising is based on what is happening in the active thread. OpenAI says signals may include the subject and intent of the current conversation, general location or language, and information about the ad itself: its landing page, headline, copy, advertiser-provided context hints, and targeting selections. If someone is discussing a home-office setup, the system may look for an offer related to that immediate goal.

Personalized advertising may additionally use selected signals from a person’s broader ChatGPT experience, including past chats, memory, ad history, and ad topics when the relevant settings are enabled. OpenAI says actions such as hiding or clicking an ad are not themselves added to ChatGPT memory. A conversation created after a user deliberately shares an ad with ChatGPT may still be referenced later if chat-history referencing is enabled.

The European Economic Area and Switzerland have a crucial launch restriction: personalized ads are not initially available. That does not make Free and Go ad-free. Contextual ads can still be selected from the current thread. Turning personalization off in another supported region creates a similar distinction: other chats, ad history, and topics are not used for selection, but the context of the active conversation can still be considered.

How to find Ad Controls and change the settings

If ads are available for the account, OpenAI directs users to Settings → Ad Controls. The Help Center also provides a direct control link for the web and iOS experience. If the section is missing, update the mobile app first. During a staged launch, the controls may not appear for every eligible account immediately.

1. Open the ChatGPT profile menu and select Settings.
2. Look for Ad Controls.
3. Check whether controls for ad personalization and the use of past chats and memory are available.
4. Disable personalization if preferred. Remember that this does not remove contextual ads.
5. Review ad history and topics, and delete advertising data if that option is available. OpenAI says full deletion from its systems can take up to 30 days.
6. Use the three-dot menu on an individual ad to hide it, report it, or see why it was shown.

There are several ways to remove ads entirely. Plus and Pro remain ad-free. OpenAI also describes a no-ads Free experience with lower message limits and reduced access to some tools, potentially including image generation or deep research. Temporary Chat does not show ads, but it is not a substitute for normal history and memory when someone wants an ongoing workspace.

What advertisers are getting

At the start of the European expansion, OpenAI plans to onboard advertisers through its Ads Solutions team, agency partners, and technology partners. The company says self-service buying through Ads Manager will follow later this summer. A small business should not assume that a complete dashboard and identical features will be available in all 31 markets on day one.

OpenAI’s launch announcement names geo-targeting, custom audiences, conversion optimization, the OpenAI Pixel, a Conversions API, and third-party measurement integrations. A separate advertiser guide describes CPM and CPC campaigns and reporting for impressions, clicks, spend, click-through rate, average CPM/CPC, and conversions.

Those capabilities should not be interpreted as a brand receiving the text of a private conversation. OpenAI says delivery decisions remain within its ads system, while advertisers receive aggregated reporting and campaign measurement data configured for their campaigns. The exact features, consent requirements, and reporting depth available to a specific European advertiser at launch must still be checked in that advertiser’s dashboard and current documentation.

A practical checklist for ChatGPT users

• Check the plan. Free and Go may contain ads; Plus, Pro, and organization plans should not.
• Update the app. An older build may not display the new controls.
• Do not confuse personalization with removal. Contextual ads can remain after personalization is disabled.
• Read the sponsorship label. A model answer and a paid unit have different status.
• Do not paste secrets merely because advertisers cannot see chats. Passwords, API keys, medical records, and client data should still be withheld unless necessary; use our AI privacy checklist before uploading sensitive material.
• Set rules for work accounts. Plan selection, permitted data, and the use of sponsored links can be addressed in an AI use policy.
• Verify the seller. Relevance is not evidence that an offer is safe, high quality, or the best price.

What the announcement still does not tell us

The official pages describe principles, not every part of the lived experience. There is not yet enough independent evidence to judge ad frequency in long threads, visual separation across screen sizes, contextual matching quality in different languages, or the effect of sponsored units on user decisions. It is also unclear when each European market will receive self-service Ads Manager and whether formats will be identical on the web, Android, and iOS.

Policies can also change as the test develops. OpenAI currently says ads are not eligible to appear near personal health, mental health, or political conversations, that political advertising is not allowed, and that accounts identified as belonging to minors remain ad-free. These are meaningful commitments, but their real-world implementation should be evaluated separately from the written policy.

Conclusion

The European launch of ChatGPT Ads primarily changes the Free and Go experience, not the mechanism that generates an answer. Sponsored units are supposed to remain separate, advertisers do not see private chats according to OpenAI, and personalization is initially unavailable in the EEA and Switzerland. The current conversation can nevertheless still inform contextual ad selection.

The sensible response is to check the plan and Ad Controls, understand that disabling personalization is not the same as going ad-free, and treat every sponsored unit as a paid claim that needs verification. Because OpenAI announced rollout for the week of August 24–30 and it may be staged, firm judgments about the European experience should wait until the feature reaches real accounts across the supported markets.