ChatGPT Ads Just Launched in Europe: What to Know

ChatGPT Ads went live in Europe on August 24, 2026. Thirty-one countries, one Monday, and most small businesses that hadn't been tracking OpenAI's ad rollout schedule got no warning at all. Here's the part almost nobody's headline led with:

ChatGPT Ads went live in Europe on August 24, 2026. Thirty-one countries, one Monday, and most small businesses that hadn't been tracking OpenAI's ad rollout schedule got no warning at all.

Here's the part almost nobody's headline led with: your customers in Berlin or Vienna started seeing ads inside ChatGPT that day. You, the business trying to buy one, mostly can't yet.

That gap between "live for users" and "open for advertisers" is the actual story if you sell to anyone in the EU, the EEA, or Switzerland. We already broke down what ChatGPT Ads mean for US small businesses when OpenAI expanded there in August. This is the European version, and the rules are different enough that a straight copy-paste of that advice would get you in trouble.

OpenAI opened 31 markets in one push

The rollout covers all 27 EU member states plus Iceland, Liechtenstein, Norway, and Switzerland, according to OpenAI's own announcement and confirmed by Euronews. That's Germany, Austria, France, Spain, Italy, the Netherlands, Sweden, Norway, Denmark, and 22 more, all at once. It takes OpenAI's ad business to 40 total countries, six months after the product launched as a US-only pilot in February 2026.

Ads only show to people on the Free tier and the paid Go tier, which runs about €8 a month in Europe. Plus, Pro, Business, Enterprise, and Education subscribers stay ad-free, the same split OpenAI has used everywhere else it has launched ads.

There's a size threshold worth knowing about too. Euronews' coverage of the launch flags that the European Commission can designate ChatGPT's search function as a "very large online search engine" under the Digital Services Act once it clears 45 million monthly EU users. Cross that line and OpenAI takes on stricter obligations: risk assessments, and a public advertising repository showing who bought what, at whom, and why. It hasn't crossed that line yet. It's close enough that the rules governing this ad product could shift again within a year of launch.

The catch: ads went live before advertisers could book them

Users didn't need to opt in to anything. Advertisers did, and mostly still can't.

At launch, there's no self-serve signup for a business in Munich or Zurich. Bookings run through OpenAI's own Ads Solutions team, or through agency and technology partners: Publicis, WPP, Omnicom, MediaPlus, Havas, and Dentsu are named in OpenAI's rollout. Self-serve access through Ads Manager, the same tool US advertisers already use, is expected "later this summer." No firm date attached to that.

GDPR is why this rollout looks different from a US launch

OpenAI didn't copy its US ad model into Europe unchanged. Contextual ads, the default, run on the conversation's current topic, an approximate city-level location, device type, time of day, and language. That processing happens under GDPR's legitimate interest basis, meaning OpenAI doesn't have to ask permission first, per reporting from Digiday.

Personalized ads are a separate track. Those use a user's account history and require explicit opt-in consent, which the person can withdraw at any time through account settings. Advertisers only ever receive aggregate metrics, views and clicks, never individual chat content, names, or precise locations.

Every platform says it protects user privacy. The difference here is OpenAI built the consent flow before regulators forced it to, not after a fine. That's a real distinction, and it's also exactly the kind of detail that changes nothing about your actual ad performance. Test the channel on its results, not its privacy copy.

— our team running paid social for DACH clients

Running AI-made creative here stacks two disclosure rules, not one

This is the part a generic rewrite of the launch news won't tell you. If your ad creative for this channel is itself AI-generated, an image from Midjourney, a video from Sora, copy drafted by an LLM, you're not just navigating GDPR's ad-consent rules. You're also inside the EU AI Act's transparency regime, which started enforcement on August 2, 2026 and requires machine-readable marking on AI-generated content.

Two separate laws, two separate obligations, one piece of creative. We've already covered whether you're legally required to label AI-generated ads in more depth, and the short version holds here: the EU AI Act doesn't care which platform you're advertising on. It cares whether the content was machine-made. A DACH business running AI-native creative through ChatGPT Ads is now the party responsible for both the platform's ad rules and the content's own disclosure requirement, and most of the launch coverage this week skipped that stacking entirely.

What the ad tooling actually looks like once you're in

Worth knowing before you evaluate whether this channel fits your budget at all. OpenAI has already built out more than a basic "boost this chat" button. Beyond simple cost-per-click and cost-per-impression bidding, the platform supports conversion optimization, meaning you can tell it to chase purchases or signups instead of raw clicks, along with geo-targeting down to a market level and custom audience lists you upload yourself.

Measurement runs through the OpenAI Pixel and a Conversions API, the same pattern Meta and Google use, plus support for third-party measurement partners if your team already reports through one. None of that is unusual for 2026. What's unusual is that this tooling exists in a market where most businesses still can't open an account. The infrastructure is ready. The door isn't.

Where ChatGPT Ads fit next to Meta and Google right now

Treat this as a channel to test in Q4, not a channel to bet the quarter on. It's day one in Europe. There's no mature bidding data, no benchmark CPCs for German or Austrian audiences, and limited access even if you wanted to spend today.

Compare that to where your budget already lives. Meta just shipped its own AI assistant that reads your ad account and tells you what to fix, a live tool you can use this week. Google's AI Max system is reshaping Search campaigns on a forced timeline of its own. Both are established, both have real conversion data behind them, and both are worth more of this quarter's attention than a channel still working out advertiser access in your country.

There are hard limits worth knowing before you plan around this channel at all. OpenAI won't run ads near health, mental health, or political content, and political advertising is banned outright. Ads are excluded from temporary chats, the Atlas browser, and any account OpenAI believes belongs to a minor. That's a narrower inventory than Meta or Google offer, and it will stay narrow for a while.

What to actually do about it this quarter

Register for updates at openai.com/advertisers so you're not finding out about self-serve access from a press cycle. If you already work with an agency partner on OpenAI's list, ask them directly whether they can book European inventory yet, because the answer varies by country and changes weekly.

If you don't have that relationship, don't build a Q3 plan around a channel you can't book. Spend the next few weeks somewhere you can already measure results: tighten targeting on the campaigns already running through Meta or Google, and use the gap to get your AI-content labeling process sorted before it becomes urgent. A business that walks into ChatGPT Ads with compliant creative already built has a real head start over one still figuring out disclosure rules mid-campaign.

Get your creative production in order regardless. Whichever channel opens first, Meta's assistant this week, ChatGPT Ads next month, the bottleneck is never going to be the platform. It's having tested creative ready to run the day access lands, built with both GDPR and the AI Act's labeling rules already handled instead of bolted on after a legal team flags it.

That's the actual advantage of watching a channel before it's fully open. You're not racing to figure out the rules while your first campaign is live. You've already had the argument about what the creative should say, gotten the disclosure language right, and built a handful of variants, so the day a self-serve button appears in your country, you spend the first week testing instead of the first month scrambling.