While European tech institutions and players debate digital sovereignty, strategic dependencies and technological autonomy, American BigTech is advancing on another terrain. Less discursive, more operational, they secure energy capacities, internalize calculation, organize the security of critical systems, structure payment flows and impose architectures which, through use, end up establishing themselves as standards.
If everyone is aware that sovereignty is no longer a matter of incantation alone, it can only exist in action. A discrepancy regularly highlighted, particularly when we compare European public discourses with the reality of cloud, energy and infrastructure capacities actually available.
Contemporary sovereignty is measured less by the control of a territory than by the capacity to guarantee the continuity of vital functions. Available and predictable energy to power AI data centers, large-scale computing capacity planned over long industrial cycles, operational security, circulation of value, access to information: these dimensions today form the backbone of economic and technological power.
Digital sovereignty must therefore be read as a problem of architecture, functional dependencies and continuity of service, well beyond an abstract institutional framework.
An actor becomes quasi-sovereign when it is able to plan these functions over the long term, internalize their critical dependencies and negotiate their deployment with public authorities (location of data centers, energy connections, taxation, regulatory deadlines). Big Tech understands this well: without seeking to replace States, they occupy an intermediate space in which infrastructure precedes rules.
Energy: the primary condition of AI
The rise of artificial intelligence has put energy back at the center of the game. Large-scale models require stable power, contracted over time horizons of ten to twenty years. This constraint explains why Meta, Google, Microsoft and Amazon are increasing the number of long-term agreements with electricity producers and supporting low-carbon projects, including nuclear ones.
This is not a simple issue of environmental optimization, but of securing supply, comparable to that of heavy industries. Energy is becoming an internalized strategic asset, just like data centers. Where the cloud relied on relative abundance, AI now imposes structured management of scarcity.
Compute: the return of industry
The promise of an immaterial digital world has dissipated. It leaves room for intensive computing, marked by high capex, complex supply chains, strong material dependencies and explicit territorial arbitrages. AI data centers operate like factories, with requirements for resilience, efficiency and continuity reminiscent of energy or telecoms infrastructures.
This reality explains the rise in power of multi-year strategies and vertical integration of computing. Platforms no longer just buy capacity: they design it, finance it and operate it.
Security: a function that has become systemic
As digital services become mission critical, security emerges as a core function. Cybersecurity, threat detection, service continuity, identity protection: major platforms now operate security systems on a global scale.
This increase in power first responds to a functional requirement: guaranteeing the reliability of systems on which businesses, administrations and citizens depend. However, it creates an asymmetry, with transparency obligations varying greatly depending on the jurisdiction. Here, sovereignty is expressed by the capacity to act in real time.
Control of value flows
Economic sovereignty requires transaction. By structuring payment, authentication and application distribution ecosystems, Big Tech controls the rails through which value circulates. Integrated wallets, in-app payments, rules for access to marketplaces are all mechanisms for deciding who can sell, pay or monetize.
Standards and architectures: governing by implementation
Technical standards shape uses before the law regulates them. APIs, protocols, identity formats, security standards are established through massive adoption, then become de facto references. The law then intervenes, most often to adjust an already stabilized state of the world.
A de facto economic diplomacy
Implementation of data centers, submarine cables, energy negotiations, taxation, regulation of AI: States have been dealing for several years with Big Tech as structuring players of their attractiveness and economic security. Germany, the United Kingdom and, of course, France are increasing their seduction operations with the leaders of these companies.
It remains to maintain a diplomacy which does not give up on subjects where the game is not completely won by the United States, without falling into a demonstration of excessive firmness on areas which do not lend themselves to it.
Europe facing the gap between discourse and infrastructure
Europe has multiplied texts, strategies and principles, sometimes too early, sometimes too late. In a fast-paced environment, pace becomes a deciding factor.
Three trajectories are emerging: regulate to contain, at the risk of lasting dependence; build own capacities, at the cost of heavy investments and assumed industrial choices; or hybridize, by combining interoperability, minimum critical capabilities and negotiated partnerships.
No option is exclusive. However, all of them require recognizing that sovereignty is no longer limited to its historical definition. The change is neither total nor irreversible, but it requires a new lucidity and a more collective approach, which is far from being an easy task.
Finance, the decisive power factor
Ultimately, the financial capacity of Big Tech remains the decisive lever of their quasi-sovereignty. Massive balance sheets, access to long-term debt and the ability to make critical investments over several decades are all assets that give them a power of action that neither the political cycle nor the European budgetary framework can currently fully guarantee.
