Who holds the knowledge within the company?
Artificial intelligence has become part of everyday work life. It summarizes documents, answers inquiries, and streamlines processes. The more useful AI becomes, the more important the question becomes: Who retains control over the data, knowledge, and decisions within the company?
Many companies today use models from the U.S. via the cloud. This approach is fast, powerful, and easy to integrate. For simple texts or public information, this approach often makes sense.
The situation is different when internal data is involved. Purchase prices, technical parameters, customer files, or production know-how are not neutral data. They often form the core of a competitive advantage. The Swiss Data Protection Act also applies to AI-supported data processing. Providers and users must be transparent about how data is processed and whether it is used for the further development of a system.
Knowledge Becomes Critical Infrastructure
AI only realizes its full potential when given context. It must understand how processes, products, and information are interconnected. Only then can it assist with decision-making, organize documents, or provide targeted support to employees.
This is precisely where the strategic significance lies. The more accurately a system reflects the company, the more valuable it becomes. The question, therefore, is not merely whether data is stored in a Swiss or European data center. It is also crucial to determine which jurisdiction the operator is subject to, who can gain access, and whether information is used to train additional models.
Three paths lead to AI
Companies can use powerful U.S. models via European infrastructure. They can rely on European providers such as the French Mistral. Or they can run open models on their own or local infrastructure.
Each option comes at a cost. Cloud services offer speed and scalability. In-house operation provides the most extensive control but requires expertise, computing power, and clear operational processes. Many Swiss companies are therefore opting for hybrid models. Sensitive applications remain in their own data centers or with local providers, while less critical tasks run in the public cloud.
Europe Is Becoming an Alternative Hub
European and Swiss offerings are gaining traction. Not because they outperform every U.S. model in every discipline, but because they give companies more freedom of choice regarding location, contract terms, and data flow.
In Zug, entrepreneur Yves Zumbühl aims to fill precisely this gap with Botts.ai. The platform is designed to enable companies to integrate AI models hosted in Europe or Switzerland into their existing systems. This example illustrates where the market is headed. AI is evolving from a single chat window to an enterprise-wide infrastructure.
It’s not origin that matters, but control
Having data located in Switzerland is not an automatic seal of quality. Even a service with servers in Switzerland can be subject to foreign law. Conversely, U.S. models are not inherently prohibited. What matters are data classification, contractual foundations, technical safeguards, and a clear rule governing which AI is permitted to access which information.
The next AI decision is therefore not purely an IT issue. It is a leadership decision. Those who clarify today where their data is processed lay the foundation for speed with control.