Ads Coming to ChatGPT Free and Go in Europe "Later This Month"

Set Trending Topics as a preferred source on Google.

Roughly six months after the US launch, OpenAI has started emailing users in the European Economic Area about a change to its privacy terms. The sender is OpenAI Ireland Limited, based in Dublin, which under the EU privacy policy acts as the controller for processing the personal data of users in the EEA and Switzerland. That makes the Irish Data Protection Commission the competent supervisory authority.

The core of the notice: ads may start appearing on the Free and Go plans “later this month”. OpenAI does not name a specific date. The Plus, Pro, Enterprise, Business and Education plans remain ad-free. The new version of the privacy policy is currently available as a preview and takes effect, according to OpenAI, “later this month”.

So ads are coming to the free tier — but also to the low-cost paid “Go” plan at €7.99 per month. How well they work from an advertiser’s point of view is an open question; in the US, ChatGPT advertising has not been a runaway success so far. Competitors such as Perplexity and Anthropic (Claude) run no ads in their chatbots, and advertising king Google has so far kept Gemini ad-free.

Contextual at launch, not personalised

For the EU, OpenAI is taking a staged approach. At launch, ads are explicitly not personalised. According to the email and the privacy policy, selection draws on:

• the current topic of the ongoing ChatGPT conversation,
• approximate location,
• device type,
• time of day and language.

Explicitly not used at launch are past chats and stored memories. As the legal basis for this generic advertising, OpenAI cites legitimate interest under the GDPR — unlike personalised advertising, which is based on consent.

Personalisation is meant to follow in a second step, and only after users are actively asked: before the company begins personalising ads, the email says, it will first ask users to choose whether they want to see more relevant, personalised ads. The help centre article on ads in ChatGPT confirms this: personalised ads are not initially available in the EEA or Switzerland.

For those who later opt in, OpenAI says the selection will additionally draw on the current chat thread including personalised model responses, the user’s own interactions with ads (such as ads they hide or click), and past chats and memories. The relevant settings are to be found under Settings > Ad controls, where ads data can also be deleted — OpenAI states it can take up to 30 days for that data to be removed from its servers.

Where ads will not appear

OpenAI has defined several exclusions: ads are not eligible to appear near sensitive or regulated topics such as personal health, mental health or politics. Political advertising is currently not permitted at all. There are also no ads in temporary chats, in the ChatGPT Atlas browser, or in accounts OpenAI believes belong to minors — determined in part through age prediction that factors in stated age, account age, times of use and the general topics a person discusses.

Ads are placed below a response, labelled as “sponsored” and visually separated. OpenAI stresses that ads do not influence the model’s answers, and that advertisers receive only aggregate metrics such as views or clicks — no chats, no memories, no names, email addresses, IP addresses or precise locations.

What other personal data OpenAI collects

The new privacy policy lists a far broader set of data than the email suggests. Among the categories collected:

• Account information: name, contact details, login credentials, date of birth, payment information and transaction history, optionally a profile picture and username.
• User content: prompts, uploaded files, images, audio and video, and data from connected services (connectors).
• Contact data: anyone who links their address book transfers contact data from their device to OpenAI — including data on people who do not use the service themselves.
• Log and usage data: IP address, browser type, timestamps, time zone, country, user agent, features used. For Free and Go users, “advertising history and interests” is explicitly added. Browsing data may also be collected when the in-app browser is used.
• Device and location information: device identifiers, operating system and approximate location; more precise GPS data only optionally.
• Data from other sources: the policy states that OpenAI receives information “from advertisers and other data partners” — information about purchases you have made is given as an example.

All of this is processed on servers outside the EEA, in particular in the United States, relying on adequacy decisions and standard contractual clauses.

Ad-free only with strings attached

Alongside the paid subscriptions, OpenAI offers an ad-free variant of the free plan. It comes with lower usage limits and reduced feature access, however — fewer messages per day and no access to tools such as image generation or deep research.

That is precisely the point likely to become relevant under data protection law. In its Opinion 08/2024 on “consent or pay” models, the European Data Protection Board held that platforms must offer a genuine, equivalent alternative without behavioural advertising. Whether a free variant with reduced limits meets that standard is an open question — the trade publication PPC Land points to ongoing proceedings against similar models run by European media companies. Most recently, at any rate, OpenAI gave free users unlimited chatting with GPT-5.6.

The road to Europe

The advertising test started in the US on 9 February 2026, followed by Canada, Australia and New Zealand. On 6 June 2026, ChatGPT Ads went live in the United Kingdom — the first European market, though outside the EU GDPR. According to an update from OpenAI dated 11 August 2026, ads have since rolled out in Mexico, Brazil, Japan and South Korea as well; further markets are to follow this year.

When advertisers in Austria and Germany will be able to book campaigns remains unclear: so far, the user-facing rollout and the opening of the ad market have run on different clocks in every market. Businesses can register for updates at openai.com/advertisers.

A second German office in Berlin

That OpenAI is taking the European market more seriously commercially is also visible beyond its advertising products. Just one day before the privacy emails went out, CEO Sam Altman announced a second German location in an interview with Handelsblatt: the company has decided to open a second office in Germany, in Berlin. He gave no opening date, address or team size — but said operations should begin “pretty soon”. OpenAI announced its first German office in early 2025 in Munich and opened it in May of that year.

The move is justified by the importance of the market: according to the company, Germany is its largest European ChatGPT market by paying subscribers and ranks among the top three worldwide. Altman also renewed his push for data centre capacity to be built in Germany, though there are still no concrete commitments on that front.

For the advertising market, the question of location is far from incidental: when OpenAI first posted advertising roles at scale in May 2026, the listings covered San Francisco, New York, London, Tokyo, Seoul, Sydney and São Paulo — continental European locations were missing. Since July, according to industry reports, the company has been building teams for France, Germany, Ireland and Singapore. A Berlin office fits the pattern: the DACH region is shifting from a pure user market to a sales market.