Digital sovereignty

Mr. Stürmer, you have been working on the topic of digital sovereignty since the 2000s. How did this come about?

Matthias Stürmer: Well, as we all know, digital sovereignty is realized using open source technologies. And I've actually been involved with this topic of open source software since I studied business informatics at the University of Bern in the early 2000s. I first went to a Linux Day in Germany in 2003 and was immediately fascinated by this dedicated community, by the top IT experts who put their heart and soul and vast tech know-how into the development of the free operating system Linux and other open source programs on a voluntary basis.

What do you understand by digital sovereignty today?

As part of a study on digital sovereignty for the Swiss Federal Department of Foreign Affairs, I analyzed over 200 academic sources last year. I discovered that there are very different perspectives and approaches to the topic: Some see digital sovereignty as an abstract, theoretical form of state independence and have written great legal and historical treatises on the subject. I, on the other hand, subscribe to the understanding of digital sovereignty as presented at the German Digital Summit 2018. This widespread definition sees the benefits of digital sovereignty in the technological autonomy of a country and its administrations and companies. Specifically, organizations should have a choice of IT companies that operate and further develop their IT systems. Digital sovereignty also means that no one can access an organization's data without explicit permission.

To what extent is the topic established in research and teaching and what gaps do you see?

The topic of "digital sovereignty" has not even arrived in education yet, otherwise we wouldn't have these major manufacturer dependencies today. At CIO level, the paradigm "Nobody ever got fired for buying Microsoft" still prevails. That's why I see a need for lectures such as "requirements engineering" and IT procurement training to consistently take "digital sovereignty" into account as an important, non-functional requirement in the evaluation and introduction of IT systems.

In research, the topic of vendor lock-in and path dependencies has been known since the 1990s, but for a long time only a few researchers have dealt with this exciting topic. In IT, dependencies are virtually omnipresent: every time a system is introduced, the integration of interfaces and the training of employees makes you dependent on a software solution. However, this system dependency is not the problem at all; in fact, it is the aim of the exercise. Only when an application is well integrated with existing systems can data be exchanged cleanly and thus fulfill the once-only principle. And only when employees know a program well can they work efficiently with it.

The problem is dependency on the manufacturers. It arises when there is only one company that can update and further develop the software. Then, for better or worse, we are at the mercy of corporate interests.

Sovereignty is a particularly important part of Switzerland's identity. To what extent is digital sovereignty an issue in the public debate and to what extent is politics concerned with it?

The political demand is not actually new, as Jonas Fricker had already submitted the interpellation "Digital sovereignty of the Swiss Federal Administration" in 2017. Little or nothing has happened, and I find it astonishing that we in Switzerland are only now seriously entering the debate and thinking about greater digital independence. Apparently it took a President Trump to startle Switzerland. At least politicians have now reacted. This year and last year, a dozen initiatives have been submitted from the left and right calling for an increase in digital sovereignty and concrete measures to implement it. Now it is up to the Federal Council and the Federal Administration.

What are the biggest opportunities and the biggest challenges on the road to greater digital sovereignty?

The biggest challenge is the reduction of vendor lock-in, but also the prevention of new, additional vendor dependencies. If you think about how time-consuming the introduction of an ERP solution is, for example, how much business logic is contained in such an application and how much an organization has adapted its processes accordingly, you probably never want to leave it again. The situational dependency must be accepted for the time being. It becomes dangerous when new requirements lead to further dependencies due to convenience ("We already have a current contract.") or incorrect investment logic ("We've already invested so much anyway.").

The opportunities of digital sovereignty must be seen in the long term. For example, public administrations will have to make fewer IT freelancers if they rely on open technologies. Competition can then play a better role. This creates choice and reduces costs. Another major opportunity of digital sovereignty is the added value that can be increased by local IT companies.

The topic of artificial intelligence has been booming for several years now. What new aspects of digital sovereignty are associated with this?

AI is guaranteed to change many things in our society and working world, and this trend is here to stay. We will see great progress in the coming years, especially in agent-based AI systems and the spread of new AI models. Interestingly, I see many parallels here with the development of the software industry 20 years ago: In the new world of Large Language Models (LLM), for example, there is also a hot discussion and parallel trends regarding proprietary versus open AI variants. Millions of machine learning models are already freely available on Hugging Face and form the basis for the many innovations. The so-called "open source AI" approach can also be pursued here.

With AI, it is also crucial how the LLMs have been trained, i.e. what data they are based on and what adjustments ("alignment" or "fine-tuning") have been made. We don't want AI platforms that reflect corporate interests or the political positions of other countries; we want objective chatbots without bias.

To what extent do Swiss companies and administrations have any chance of creating or maintaining digital sovereignty for themselves?

It is a myth to believe that the state and Swiss companies are forever and ever dependent on foreign tech companies. We can very well build and operate cloud platforms, software applications and AI models based on open source technologies ourselves. With the "Network SDS - Sovereign Digital Switzerland" founded in July 2025, we see dozens of Swiss companies offering cloud services for public authorities and the Swiss economy. On the "OSS Directory", there are hundreds of open source products and many competent providers who can help implement digitally sovereign solutions. And in the field of AI, our top universities ETH Zurich and EPF Lausanne, together with the Supercomputing Center CSCS in Lugano, have developed a modern, completely transparent open source large-scale language model. This clearly shows that we are perfectly capable of implementing digital sovereignty. Often, we simply lack visibility, networking and confidence that we can do it. That is why we want to create more visibility for digitally sovereign solutions from courageous organizations with the SDS network and the OSS Directory.

In Switzerland, we are lucky enough to be at the forefront of research and teaching. What do you wish for yourself and our society in the coming years?

Switzerland has ideal conditions: A high level of education with world-renowned universities, top IT infrastructures with numerous data centers and a unique AI supercomputer in Ticino, a stable power supply, a high level of political continuity, hardly any debt, a strong currency and an excellent image abroad for security and reliability. Now we need to catch up where other countries are doing better: We need more courage and more cooperation.

Author

Pascal Sieber is a transformation consultant and Chairman of the Board of Sieber & Partners. He is the founder and organizer of the annual CNO Panel. www.sieberpartners.com/cno-panel-aktuell