Technological Sovereignty

Almost everything that runs day to day in a country depends on technology from abroad. The electricity for factories comes from power plants, the software to control them often comes from the US. The chips in cars, phones, and washing machines nearly all come from Asia. Technological sovereignty describes how strongly a country can control such dependencies. It does not mean building everything oneself. It means the ability to remain capable of acting in an emergency: because one has alternatives, because one could produce it oneself, or because one works together with many different partners.

When a supply stop paralyzes a country

The term became popular in Europe mainly through two events. During the Covid pandemic, semiconductors—the small computing components in almost every device—were suddenly missing. European car factories came to a standstill because a few chips worth just a few euros could not be delivered. Afterward, the war in Ukraine showed how quickly an energy dependency can become a political weapon.

With software, the dependency is less visible, but no smaller. A large share of European public authorities and companies store their data with three American providers: Amazon, Microsoft, and Google. Switching is expensive and takes years. Whoever is already in negotiates from a weak position. Experts call this the lock-in effect.

With AI, there is an additional point. Whoever builds the large language models also decides what answers they give and which topics they avoid. These decisions follow the rules and values of the country of origin. Europe is therefore discussing whether it needs its own models, ones subject to European law.

The four layers of dependency

The question can be broken down into layers. At the very bottom lies hardware: chips, machines, raw materials. Above that lies infrastructure, meaning data centers, networks, and cloud services. Then comes software, from the operating system to the AI model. At the very top are the data and the rules by which they are processed.

Sovereignty does not mean being independent at every layer. That would be unaffordable. No country in the world builds the entire chain alone, not even the US or China. Realistically, it is about targeted strengths. Europe, for example, leads in machines for chip manufacturing: the Dutch company ASML builds equipment that no one else in the world can produce. This single point in the chain gives Europe negotiating power.

A second lever is open software. If a program’s source code is publicly viewable, anyone can inspect, copy, and continue operating it. The program then belongs to no single company. That is why many administrations deliberately rely on such open solutions. A third lever is simply diversity: two providers instead of one, even if that costs a bit more.

From chip factories to the school cloud

In the news, the term usually appears alongside billion-euro sums. The European Chips Act is meant to significantly increase Europe’s share of global chip production. Countries lure manufacturers with subsidies, such as Intel to Magdeburg or TSMC to Dresden. Whether such projects work is disputed, since they are expensive and take a decade.

But the term also exists on a smaller scale. Some federal states argue over whether schools may use Microsoft programs or need their own solution. Hospitals ask where patient data is stored. Companies examine whether they want to entrust their design plans to a foreign AI.

A common misconception: sovereignty is the same as protectionism, meaning sealing off one’s own market. It is not. Whoever allows only domestic technology often ends up with worse and more expensive products. The goal is freedom of choice, not isolation. Whether Europe manages to strike this balance is one of the central economic policy questions of the coming years.