ChatGPT ads just landed in Europe. But not the version everyone else is getting.
Ads are now available across 31 European markets. OpenAI has launched self-service Ads Manager access, conversion optimisation, custom audiences and measurement tools. It says tens of thousands of marketers are already advertising on ChatGPT globally.
But there is an important European twist.
Personalised ChatGPT ads are not initially available in the European Economic Area or Switzerland.
That makes Europe an interesting test case for what AI advertising looks like when conversational intent is powerful, but broader behavioural personalisation is constrained.
And it is happening just as ChatGPT comes under a much tougher EU regulatory regime.
First, ChatGPT is now officially search
On 31 August, the European Commission designated ChatGPT a Very Large Online Search Engine, or VLOSE, under the Digital Services Act.
The Digital Services Act, or DSA, is the EU framework governing online platforms and search engines.
ChatGPT crossed the threshold after reporting more than 45 million monthly users in the EU. OpenAI reported around 159 million.
The designation puts ChatGPT into the same regulatory category as Google Search and Bing.
That matters for marketers because it formalises something already happening in user behaviour.
People increasingly use ChatGPT to research products, compare providers and make decisions.
And OpenAI is now building an advertising business directly around those moments.
AI advertising starts with intent, not keywords
Search advertising has always worked because queries reveal intent. Someone typing “best payment orchestration platform Europe” is worth more to an infrastructure company than someone who just matches a demographic.
ChatGPT goes further. Instead of three keywords, people describe the actual problem. A marketplace expanding into Germany and France, asking what to consider in a payment provider that supports local methods. That single conversation hands an advertising system far more context than any keyword ever could.
OpenAI says ChatGPT Ads use the context of the current conversation to decide what to show. Advertisers can also submit “context hints,” topics or conversations where their product might be relevant.
Not exact-match keywords. Conversational relevance.
For fintech, that matters. Payments, compliance, banking infrastructure, none of it is a simple purchase. The buyer’s actual problem tells you more than any query string.
Europe gets a narrower version
Outside Europe, personalisation goes deeper. If users opt in, OpenAI can draw on signals from the wider ChatGPT experience: past conversations, and where it applies, memory.
That stacks into something powerful. Search intent, plus conversational context, plus user history.
In the EEA and Switzerland, that third layer isn’t there. Not initially, according to OpenAI’s own documentation. European users still see ads shaped by the conversation they’re having right now. Past conversations, ad history, broader topics: none of that feeds in.
To be honest, that’s the real story here. Not whether Europe gets ads. It already has them. The real question is how far personalisation can go under European rules.
Then the Privacy Center showed up
This week OpenAI started rolling out a Privacy Center inside ChatGPT. Memory, chat privacy, connected apps, data use, ad personalisation, all in one place. One section is literally called “Personalized chats & ads.”
Looks like a routine product update. It isn’t.
It gives users a clear map of the data layers sitting behind a conversation: current context, memory, past chats, location, connected apps, ad personalisation.
OpenAI is also drawing a hard line between the answer ChatGPT gives and the ads shown around it. Ads are labelled, separate, and the company says they don’t influence the answer itself.
That separation is going to matter a lot. Because there are now two distinct ways for a company to show up inside AI search.
Paid visibility and AI visibility are not the same problem
You can pay to advertise inside ChatGPT. You cannot pay ChatGPT to include your company in its organic answer. Two separate systems.
So when someone asks which European fintech infrastructure companies to consider, there are actually two marketing problems hiding in that one question.
Can your company become part of the answer? That’s AI visibility. Does ChatGPT understand what you do. Can it verify your positioning. Does it find authoritative sources discussing your company. Is your regulatory status clear. Are you filed under the right category. This is Answer Engine Optimisation, AEO, and Generative Engine Optimisation, GEO.
Can you advertise around that decision? That’s paid distribution. ChatGPT Ads already appear beneath responses. Advertisers can optimise for impressions, clicks, conversions.
These two disciplines could end up as central to marketing teams as SEO and paid search once were. They should not get confused with each other.
The DSA makes that separation matter even more
The VLOSE designation brings obligations. Independent audits. Systemic-risk assessments. Advertising transparency. Researcher access to data. Rules around recommender systems. VLOSEs also have to offer a non-profiling recommender option and keep a public ad repository.
OpenAI has until January 2027 to comply.
So two things are happening at once. OpenAI is building a sophisticated ad platform, fast. European regulators are gaining real visibility into how large search and recommendation systems actually work.
Those two forces are going to collide.
Why fintech marketers specifically need to pay attention
Financial services carry more weight here than most categories. Trust, regulation, jurisdiction, these matter far more when someone is picking a banking or payments provider than when they’re buying shoes.
If someone asks ChatGPT for a stablecoin infrastructure provider in Europe, or a payments platform suited to an EU marketplace, the answer quality depends on whether the system can establish hard facts. Licensing. Geographic availability. Regulatory status.
That leaves fintech marketers with three jobs.
Be understandable. Say plainly what your product does, who it’s for, where it operates.
Be verifiable. Back every claim with primary sources, regulatory documentation, data, credible third parties.
Be discoverable. Know how your company shows up in AI-generated answers, and know how paid AI distribution fits your acquisition strategy.
Search marketing is being rebuilt in real time
Google gave us a clean split. Organic on one side, paid search on the other.
ChatGPT is rebuilding both layers, just inside a conversation. Organic discovery becomes: are you part of the answer? Paid discovery becomes: can you reach someone while they’re deciding?
In Europe, there’s a third layer on top. What is the platform legally allowed to use to decide what that person sees?
That question alone could push the European AI advertising market down a completely different path than the US.
Which is exactly why the Privacy Center matters more than it looks. It’s an early interface for something we’re all going to be talking about for years: where personalisation ends, advertising begins, and AI search sits in between.

