A battery has no place in a debate about digital sovereignty. That’s the initial assumption—and it’s wrong. Beneath cloud contracts, network diagrams, and questions of applicable law lies a layer that we never map out because we take it for granted: the power supply. Yet this layer is undergoing a fundamental shift. As long as it relied on controllable power plants, we could ignore it. As the energy mix shifts toward intermittent sources, energy storage is becoming a full-fledged building block of infrastructure. And this building block is governed by an industrial chain as concentrated as that of semiconductors.
Let’s state this calmly, without resorting to doomsday scenarios.
Why the Battery Is a Layer of Infrastructure
All physical digital infrastructure requires a continuous power supply. A data center cannot tolerate an outage: between the moment the grid fails and the generators kick in, there is a gap of a few seconds to a few minutes that must be bridged without losing a single transaction. This role has always been filled by inverters—UPS systems—backed by batteries. It is the oldest and most commonplace function of electrochemical storage in the industry. It goes unnoticed because it simply works.
What is changing is the scale and the level of dependence. The newest facilities are replacing some of their lead-acid batteries with lithium-ion packs, which are more energy-dense and easier to manage. Operators are beginning to connect their data centers to stationary energy storage systems (BESS), which no longer serve merely as a backup buffer but as a stabilizing force in the face of a grid whose output is becoming increasingly variable. Wind and solar power cannot be controlled: they generate electricity when the wind blows and when the sun shines. For a digital service to deliver on its promise of continuity with such a mix, energy must be stored.
The same logic extends all the way to the network’s edge. Mobile antennas, edge sites, and isolated relay stations often run on batteries, sometimes operating completely off-grid at locations without a reliable connection. 5G is increasing the density of these locations. Each new antenna is a small node that depends on a battery. The continuity of digital service—from the central data center to the neighborhood antenna—therefore relies increasingly on storage capacity that we purchase, replace, and do not manufacture.
Who Controls the Production Chain
The lithium-ion battery is not a single product; it is a chain. And each link has its own geography of power.
Upstream are the raw materials. Lithium is mined primarily in Australia and South America. More than two-thirds of cobalt comes from the Democratic Republic of the Congo. Graphite, manganese, and nickel each have their dominant regions. This geological concentration is a first point of dependence, but it is not the most fundamental factor.
The decisive link is refining and processing. Extracting ore is not enough: it must be purified, transformed into electrode materials, and then into cells. That is where the concentration is greatest. China refines the majority of the world’s lithium and cobalt, produces most of the anode and cathode materials, and assembles a large portion of the cells. Owning a mine in Australia does not guarantee self-sufficiency if the next step necessarily involves a limited number of actors located elsewhere. Dependence does not stem from the raw resource itself, but from processing capacity. This is exactly the pattern we see with silicon: the ore is everywhere, but smelters are rare.
For an infrastructure operator, the consequence is clear. The battery it installs—whether it powers an inverter or a storage facility—depends on a supply chain in which no critical link is located within its territory or under its jurisdiction.
Europe’s Industrial Response
Europe has recognized the problem and attempted to address it. In 2017, the Commission launched the European Battery Alliance, billed as the “Airbus of batteries”: bringing together manufacturers, governments, and funding to build a supply chain from ore to battery cells. The goal is not merely the rhetoric of reshoring; it is to achieve production autonomy for a component deemed strategic for the automotive industry, the energy sector, and, by extension, infrastructure.
The most visible industrial arm is ACC, Automotive Cells Company, a joint venture between Stellantis, Saft, and Mercedes, with plants planned in France, Germany, and Italy. The project is underway, and the first gigafactory in Douvrin is already in production. But the trajectory reveals the real difficulties. ACC has suspended or postponed the German and Italian projects while it assesses technology and demand. Other European projects, led by players such as Northvolt, have suffered severe setbacks. The cause is not a lack of political will or public funding: it is the sheer difficulty of building a mass-scale industry in a region where the know-how for producing high-efficiency cells has been accumulating elsewhere for fifteen years. Factories can be funded quickly. But you can’t recreate a learning curve by decree.
The result, at this stage, is a gap between stated ambition and installed capacity. Europe has projects, a few sites in production, and a dependence that remains largely intact on solar cells and, above all, on upstream processed materials. The industry exists on paper and, in part, on the ground. It does not yet meet the demand.
What This Means for an Operator
The lesson is familiar to anyone who has followed the cloud debate. Sovereignty isn’t determined by the flag on the finished product; it’s determined along the entire supply chain. A battery assembled in Europe may depend, in the previous stage, on materials refined in a single country. Asking where the battery is assembled is not enough, any more than asking where the data is hosted. The right question concerns the links in the chain that we cannot see.
In practical terms, an infrastructure operator’s exposure isn’t just energy-related in the sense of electricity prices. It’s material. Its service continuity relies on batteries whose supply, replacement, and cost depend on a concentrated supply chain that’s vulnerable to trade tensions. A disruption in the supply of battery cells, an export restriction on a critical material, or a spike in the price of refined lithium all affect the ability to maintain and expand the infrastructure—and thus the service.
There is no SecNumCloud certification for batteries, and there won’t be one. But the analytical framework is the same as for the rest of the sovereignty stack. Which link does my power continuity depend on, and who controls it? Where do my backup storage and stationary storage come from, and what happens if the supply chain is disrupted? Do I have a secondary source, a reversible supply, or a buffer stock? These are questions of procurement and resilience, not communication. They arise alongside questions of applicable law, because a service that is sovereign on paper but shuts down due to a lack of reliable power is no longer sovereign at all.
The battery is not the heart of the debate on digital sovereignty. It is its foundation. We always end up coming back to it, because nothing runs without power, and power, from now on, can be stored.
Sources
- European Commission, European Battery Alliance, presentation and progress reports
- ACC (Automotive Cells Company), communications on the gigafactories in Douvrin, Kaiserslautern, and Termoli
- International Energy Agency, reports on supply chains for critical minerals and batteries
- U.S. Geological Survey, Mineral Commodity Summaries, lithium and cobalt
- Public data on the refining and assembly of lithium-ion cells by geographic region