# When AI Simply Gets Switched Off: Europe Needs Sovereignty but Doesn't Have It > On 12 June 2026, two of the most powerful AI models were blocked by order for all users outside the United States. No price, no technical fault, a political decision. What that means for Europe, and why we shouldn't act surprised. *Published: 2026-06-15* *Source: https://vellmerk.ai/en/blog/ki-wird-abgestellt* On 12 June 2026, a single official order was enough to shut off two of the most capable AI models in the world for all users outside the United States. No outage, no payment problem. A political decision, by letter, in a single day. In a few years we will hear it again: "No one could have seen that coming." Yes, they could. You can see it coming. Vellmerk.ai already sees it coming today. ## Two Kinds of Barriers Anyone who believes that access to leading AI is mainly a question of money is missing the signs of the times. There are two kinds of barriers, and the more dangerous one is not the financial barrier. The **monetary** barrier is the smaller one: top models become more expensive, rationed. With budgets you can work around it. You could also call it a problem you can throw money at. The **political** barrier is the real danger: blocks and allocations as an instrument of power. Whoever decides who gets access to the most capable AI holds a geopolitical lever in hand. And levers get used. The 12th of June was therefore the day a thought experiment became a precedent. ## It Is Not Purely a Hyperscaler Question Either The pattern is bigger than one model and bigger than one provider. The same logic applies to the large cloud platforms: there too, an export restriction or sanction can cut off entire services. And you only need to think of the export restrictions on AI chips toward China to see which direction this is heading. Today it is about one model with an improvement over its predecessor. Tomorrow it can be about a great deal more. ## And Europe? Watches. We have no comparable models of our own, no levers, no sovereignty. We are a customer, not an actor. And customers can be switched off. What is most troubling about this: we can see it all coming. And yet too many important people, in business as in politics, do too little to confront this risk. When it finally blows up, the same actors will step in front of the microphones, explain that it caught us off guard, and announce some emergency program to paper over the failures. ## The Good News: We Can Act Ourselves We can spare ourselves all that. Because the good news is: we do not have to wait for the grand European answer in order to become capable of acting ourselves. From multiple client projects Vellmerk.ai knows that every organization can act on its own responsibility today and position itself sovereignly, by actively reducing its dependence on individual providers and jurisdictions. The concrete path there runs through open source, European providers, open models and local, on-premise-capable deployments, not as a stopgap, but as a strategy. Why exactly this works, why no one can simply switch off a locally operated model, and what that means in practice, Vellmerk.ai has written down in detail: [Local AI: Why Independence Becomes a Strategy](/blog/lokale-ki-souveraenitaet). > "We see it coming. So let us act now, while it is still a decision and not an emergency.", Thorsten Vellmerk ## Your Next Step The question is no longer whether, but how fast. If you want to know where your organization depends on individual providers today and which workloads can be secured independently, Vellmerk.ai looks at it with you in a no-obligation initial conversation. ## About Vellmerk.ai Vellmerk.ai is an AI consultancy (Danish ApS) founded by Thorsten Vellmerk. Drawing on 20+ years of process and IT experience and several years of hands-on AI consulting, Vellmerk.ai helps SMEs and public administration adopt AI in a practical, sovereign way, from strategy to local, on-premise-ready implementation. Proven across multiple client projects. [Book an initial consultation](/en/contact). ## Frequently asked questions ### Can an AI provider really just switch off access? Yes, and that is not a theoretical scenario. Access to cloud AI hangs on contracts, export rules and political decisions you have no influence over. A changed sanctions situation, a provider decision or a new regulatory requirement can cut off a service overnight. Whoever builds a business process firmly on an external service takes on that risk as well. ### What risk arises from dependence on US cloud AI? A twofold risk arises: legal, around data transfer to third countries (the Schrems II issue), and operational, around availability. Both lie outside your control. For non-critical tasks that is acceptable. For processes that have to keep running, or for sensitive data, it is a concentration risk that should be deliberately assessed and hedged. ### How do you protect yourself against the loss of an AI service? Through independence in the right places: base critical processes on open models you can run locally, keep providers interchangeable instead of hardwiring them in, and have a fallback plan for the most important use cases. The point is not to demonise cloud AI, but to avoid being held hostage at the decisive points. ### Does independence mean giving up cloud AI entirely? No. It is about a deliberate split: cloud AI where it is cheap and unproblematic, local and sovereign solutions where data is sensitive or an outage becomes expensive. The art lies in drawing the line correctly, not in a blanket decision for or against the cloud.