Digital sovereignty: the end of the open Internet as we know it? (Part 1)

Claims to strengthen digital sovereignty have echoed in recent years, from Brussels to Addis Ababa. Sovereignty inspires policy approaches formulated by governmental and non-governmental actors, as exemplified by the India Stack and the Euro Stack. But what does sovereignty mean in an interdependent digital economy? 

In 1987, the American group R.E.M. sang about ‘the end of the world as we know it’ (a great song, worth listening). The new world born after the fall of the Berlin Wall seemed to announce the triumph of liberalism and the age of global flows. Is the current decline of the liberal order marked by the end of the open Internet as we know it? These are the questions that motivated this two-part series of blog posts. 

Part 1 places digital sovereignty in the context of the political economy. It explores why digital sovereignty switched from being ostracised to a central notion of current digital policy-making. While sovereignty was used to contest economic and rights-based inequalities during the last decade, digital sovereignty is currently under conflicting pressures. It risks being captured by securitizing narratives and by a neo-mercantilist political project. These reflections provide the canvas to connect digital sovereignty and openness in part 2.

1. Sovereignty and autonomy

Reflections on sovereignty often trace it back to the Peace of Westphalia, when the notion of State sovereignty was forged, grounded on territorial integrity, non-interference and legal equality among States. Even if we conceive sovereignty as a relational concept, co-constructed and shaped by discourse across time and culture, this historical birth remains a relevant reference point. When we seek to ‘unthink’ classic notions of sovereignty in order to define what digital sovereignty could mean, state sovereignty offers the benchmark to be deconstructed. 

For the purpose of this particular reflection, we understand sovereignty along the lines proposed by Geenens: “when speaking of sovereignty, we invoke the perspective from which a political community can consciously understand itself as an autonomous agent”. This understanding brings about important consequences. The first is that sovereignty is the manifestation of a political claim, made by conscious agents. This means that companies, for example, are not sovereign, although some of them can wield more power than some sovereign states. The second is that autonomy – understood as the capacity to steer one’s own course of action rationally, and not just as an automatic response to societal or economic inputs – is a key element. 

Much of the concerns that underpin digital sovereignty claims at present times are related to increasing the scope of autonomy. In the context of an interdependent digital economy, the desire for increased autonomy cannot realistically be translated into a search for autarky, total independence, or self-sufficiency. Autonomy seems to be rather related to the capacity of a social system – a state, for the sake of example – to choose how to respond to external stimuli (or to external ‘noise’). 

In biological systems, evolution lies in the delicate balance between minimizing external disturbance while welcoming the noise necessary to promote adaptation and evolution. In present discussions about digital sovereignty, shifting perceptions about the precise point of this balance, and about the autonomy to redefine this point if needed, are key elements hiding at the backdrop of concerns.

2. Digital sovereignty and political economy: an unfinished play in three acts

Discussions on digital sovereignty are marked by the contrast between the territorially-based operation of our political and legal systems, on the one hand, and the transborder operation of the digital economy and its underlying technological infrastructure, on the other. For the last decades, actors have tried to resolve this tension in  different ways, from the liberal rejection of the notion of digital sovereignty to the current neo-mercantilist approach, which re-erects sovereign borders along a geoeconomic projection of power. Recent shifts in US politics have accelerated this turn, but it precedes the Trump administration.   

Act I: the liberal rejection of digital sovereignty   

The liberal approach tackled the tension between territorial and transborder by removing the barriers to cross-border flows, in order to facilitate the sprawling of the techno-economic complex. According to the liberal discourse, the role of states was to provide a secure and predictable environment for the operation of transborder actors. 

The United States, in particular, engaged in intense diplomatic work during the Clinton Administration to open the path for the Information Superhighway – albeit the resistance of some Western European countries, and some vocal developing countries, such as Brazil and India, particularly during the WSIS. 

Under the bonnet, however, US public spending on technology development and innovation continued high, showing that the US has been an entrepreneurial state in the technology realm, including in the digital sector. The government committed to “spend money twice as fast as the Pentagon’s Star Wars anti-missile program” in areas that included fiber-optic communications and national computer networks. Liberalisation of the digital sector was not considered at odds with public investment at that time. Other countries took relied more heavily on attracting foreign investment and facilitating market access as  a way to develop their digital sectors.

Source: The NewYork Times, November 10, 1992. 

While the ideational patterns of interaction between public and private realms were defined within the internet governance regime (i.e. private led, multistakeholder), the binding rules were developed through the international trading system. Trade agreements were focused on removing tariff and non-tariff barriers. They successfully helped to reduce costs, cut red tape, minimize discrimination between domestic and foreign providers, and increase predictability. Foreign Direct Investment in the digital sector bloomed, as the importance of US capital in building Chinese Internet backbones exemplifies. 

In parallel, liberalisation also reduced the ‘policy space’ available to governments, including with regards to regulating digital flows. In doing so, the liberal order narrowed the range of possibilities to define the ‘sweet spot’ between unacceptable noise, and evolution-promoting disturbance. It brought about the fast-paced development of the digital economy, and ushered in the ‘age of digital interdependence’, which was expected to increase prosperity and freedom for all. Liberals celebrated interdependence as a peace-promoting feature of the international economic order.

Against the backdrop of digital optimism, digital sovereignty continued to be defended by a few countries, notably Russia and China, but it was pushed to the outskirts of policy discussions, becoming a synonym of backwardness, authoritarianism, or protectionism. Digital interdependence was seen as a natural byproduct of a globalised digital economy.

Act II: digital liberalism with a social project 

The compromise that underpinned liberalism (a narrower scope of policy options in exchange for increased prosperity, if we put it in a simplistic way) was undermined by the growth of two types of digital inequality: economic, and rights-based inequalities. 

The first of them is related to the growing inequality in the distribution of the means of production in the context of the digital economy – notably technology, infrastructure (i.e communication networks, data), and knowledge resources. This has also led to inequality when it comes to the distribution of dividends generated by the digital economy. 

In the last decade, reports by several organizations, such as the World Bank (2016), the Internet Society (2019), the World Economic Forum (2021), and UNCTAD (2021) have shown that a few countries and players are majoritarily accruing the benefits and wealth stemming from the digital economy. This growing economic concentration has led to the formulation of a postcolonial view on digital sovereignty, aimed at challenging the appropriation of digital resources by powerful actors, the dispossession of communities (i.e through data extractivism), and a process of digital colonialism.

Post-colonial lenses are enlightening in discussions on digital sovereignty, but the social and geographic contours of the digital divide do not neatly follow the North-South divide, but can be better captured by the notion of center-periphery. Center and periphery are not engaged in a linear hierarchical relation, but come together in a network. The central nodes (the ones with control or influence over the digital means of production) are capable of affecting the broader network of actors in a significant manner, creating relations of asymmetric interdependence. This helps to understand claims of digital sovereignty being formulated by developed countries and regions. Europe also fears becoming a digital colony. 

In the European Union, inequality-related concerns were framed as an asymmetry of power that led to the erosion of competition in the digital space, on the one hand, and to the undermining of individual rights and freedoms, on the other. This is the rationale that underpins the Digital Markets Act (DMA), the Digital Services Act (DSA), and the General Data Protection Regulation (GDPR), for example. From this perspective, the State (or, in this case, the Union) had to be strengthened vis-a-vis the market, since only public authorities would have the necessary tools to ‘tame’ market forces, protect the most vulnerable, and level the playing field to allow competition. This strengthening was not formulated as a challenge to liberalism, but it created a tension with the open liberal order by imposing limitations on cross-border flows, while seeking to redefine the threshold of ‘noise’ social systems were willing to accept if they wished to preserve their values and social standards . 

It is no coincidence that, in recent digital trade negotiations, the sections related to exceptions to liberalising ideas became one of the main sticking points, often opposing the US and the EU, as negotiations on data flows and privacy-related exceptions in the WTO Joint Initiative on e-commerce exemplified. 

The cracks in the liberal economic order started to become visible, but asymmetric relations were not intrinsically a problem, as long as the more dependent actor in the relationship had a margin of maneuver that allowed it to perceive itself as an autonomous agent, capable of adjusting the balance of how much noise and disturbance it would be willing to accept in its social system. The perception that this margin has shrunk is triggering discussions on autonomy and sovereignty. At the 2025 AI Action Summit, US Vice-President, J.D. Vence, conditioned the collaboration of the US with other countries in the field of AI to the adoption of regulatory regimes that “foster the creation of AI technology rather than strangle it”, something interpreted as message to the EU Commission.

Act III: the age of neo-mercantilism and the return of digital sovereignty 

Views on globalization have radically changed in the US. On the one hand, the US has made a U-turn in some of the pillars of its liberal trade policy, such as provisions on cross-border data flows. This is significant, considering the role that US-led trade agreements had on the liberalization of the digital economy across the globe. On the other hand, interdependence is mostly seen not as a feature of the ecosystem, but as a flaw that needs to be addressed. The US government claims that globalized value chains are undermining American prosperity. According to Vence, the globalized production of goods led to the flow of American knowledge and its appropriation by other players, eroding US leadership. Bringing production ‘back home’ is a way to counter these leaking flows of knowledge, which are making adversaries more competitive. 

As the US turns its back at an open liberal order – the ‘Liberation Day and the imposition of across-the-broard tariffs seems to cut the final ties of commitment – there are growing claims for increased autonomy in other countries (often framed as digital sovereignty at the level of discourse), as well as a belief that, failing to meet these claims, would lead to (existential) threat. 

Perceptions are changing drastically and fast, because the political project of liberalism is being overridden by a neo-mercantilist political agenda, which gives precedence to the state in relation to the economy. Nevertheless, in the context of this change, catering to social needs is not at the forefront of concerns, as an European ‘social-democrat’ project proposed. The precedence given to the state is seen as a necessary condition to hedge against external threats and promote (national) security. This logic manifests itself in the form of state capitalism, or as populist economic nationalism in the US, through its ‘make the US great again (MAGA) approach. 

For this reason, it would be imprecise to assume that the present US administration represents a triumph of US liberal forces epitomised by an alliance between the US government and the private (tech) sector. Although some tech companies are benefiting from a convergence of interests – AI companies provide an example – others are being confronted with the clash between their economic interests (i.e. market access, receiving higher returns from international investment) and the Government’s geopolitical and geoeconomic agendas. The (so far unsuccessful) opposition of large US companies, such as Microsoft, Amazon, Nvidia, Oracle, and the Computer & Communications Industry Association to the interim final rule that sets a framework for AI diffusion is one example of the precedence given to the interests and priorities of states over the interests of market actors.  The rule establishes export controls on semiconductors, AI model weights and on cloud-based AI as a service. The rule was introduced by former president Biden, and, so far, upheld by president Trump.

Since networks of interdependence were sewn predominantly by companies and scientific communities, the speed at which the neo-mercantilist project has managed to dislodge the liberal project and taken hold of institutions is remarkable. This has been largely facilitated by a successful securitizing discourse, which can be found in the neo-mercantilist formulation of what economic security entails.

For liberals, economic security refers to the preservation of a stable economic order (and its supporting institutions), in which free markets can operate unhindered by non-market considerations. For social-democrats, economic security means strengthening the state, so it can protect the vulnerable and promote redistribution. Neo-mercantilists give logical precedence to the goals of the state, as the only actor capable of providing the security necessary for markets to function and for societies to thrive in an anarchic and competitive world. The economy must, therefore, assist the state in fulfilling its goals. In other words, “economic security is national security”, as the US 2021 National Security Strategy affirmed. 

The neo-mercantilist approach to economic security, and the notion of national security both consider the ‘state’ as the object to be protected, facilitating the conflation at the level of discourse. This imposed an adversarial reasoning to economic policy, exemplified by expressions such as ‘trade war’, ‘tariff war’, ‘weaponization of interdependence’, ‘siege of algorithms’ and, ultimately, weaponization of the economy. 

National security is, by excellence, the terrain of military logic and existential threats, since it ultimately aims to safeguard state sovereignty against external sources of danger. This life-or-death framing provides states with the authority to push for exceptional measures, often approved without proper democratic discussion. The fact that the US is resorting to presidential Executive Orders to declare an increasing number of national ‘emergencies’ related to the digital sector –  from information and communication technologies and services to AI – shows the erosion of democratic control. Many of these EOs take legal authority from the International Emergency Economic Powers Act (IEEPA), an instrument coined within a wartime mentality.

Source: CCIA Statement on the Export Control Framework on AI Diffusion (13 January 2025)

In Europe, recent non-governmental proposals for digital sovereignty remain discursively committed to the people-centered notion of the social-democrat project. Proponents of the EuroStack aim to empower an European private sector, capable of promoting indigenous innovation that aligns with European values of transparency, accountability, and privacy. The Democratic and Ecological Digital Sovereignty Coalition argues for a strengthened state, capable of designing autonomous institutions that can provide infrastructure and essential digital services as public services or commons, enabling the mastery over the production and deployment of the digital technologies by the community. 

Nevertheless, there is a broader geo-politicisation of EU policy-making, and the shaping of ‘digital sovereignty’ by Brussels seems to be distancing itself from the concept of digital sovereignty as something obtainable on an individual level. It is becoming increasingly similar to the idea of ‘strategic autonomy’, borrowed from the defence sector and previously employed by the EU, focused on addressing vulnerabilities related to supply chain disruptions, third-country dependencies, and geopolitical risks. The direction of the EU’s political project remains unclear. 

If a neo-mercantilist view successfully takes root, exploiting economic interdependencies, and exercising coercion to achieve geopolitical goals may become the expected state of affairs. In a context of an offensive and chauvinist turn in US policy, the popular magazine The Economist, suggested a range of potential choke-points that could be adopted by the EU in retaliation. They included restricting access to the machines essential to the production of advanced semiconductors, and applying taxes and heftier fines on Internet companies, for example. In the race to discover whose economic policy decisions can cause the most pain, deeply interdependent digital value chains become an obvious choice of ‘weapon’, and digital sovereignty could strengthen the authority to deploy the arsenal.  

3. Digital autonomy as a socially-anchored and socially-driven project 

Claims for digital sovereignty must be cast against this complex and uncertain background. At the present political juncture (in which a neo-mercantilist view is increasingly becoming pervasive) claims for sovereignty risk being appropriated by a securitizing state bureaucracy. Strengthening States’ control over digital assets in a context of growing chauvinism and of democratic backsliding (across developed and developing countries) could lead to a slippery slope towards a controlling bureaucratic apparatus, which puts its geopolitical project and its quest for national security ahead of individual rights, social goals, or environmental interests.  

At the same time, sovereignty is a rallying banner, easy to communicate. It is unlikely that actors will further scrutinize the desirability of its use when the need for action seems so urgent. In this context, some actions could be put in place to mitigate risks. 

1. The decision-making and governance of policies being adopted under the banner of digital sovereignty must be socially-anchored and socially-driven. This means not only avoiding the use of regulatory instruments that do not allow due debate and political scrutiny, but also means actively creating and empowering governance arrangements tasked to ensure that sovereignty does not only benefit the state and some domestic ‘champions’. Replicating models that have created inequality (economic or rights-based ones) does not amount to a political project that benefits the autonomy of political agents.
2. Just like ‘unthinking sovereignty’ remains important, it is also urgent to rethink and reclaim ‘economic security’. The conflation between economic security and national security is deceiving. It casts the bulk of international economic relations in a Schimittian friend-enemy logic, and leads to zero-sum games. It opens the gates for securitization, by creating an (ever expanding) open-ended list of transactions that could be (potentially) relevant to national security. Such a nexus can only be convincingly established in very few cases. Economic security can be framed as cooperative rather than confrontational pursuit; a shared condition rather than a goal of individual states.
3. Contesting inequalities remains as urgent as ever. It is important, however, to operate in a multi-level playing field. While digital sovereignty remains imbued of a galvanizing post-colonial ideational force in the Global South, it is important to operate with the logic center-periphery to be able to identify potential development-promoting alliances with developed countries struggling to build their own autonomy. In addition, although the securitizing stance is less present in developing countries, the security oriented mindset which is percolating the developed world cannot be ignored, as it will be a key organising logic for technology and knowledge diffusion. 
4. There is a need to differentiate between threats to classic state sovereignty perpetrated through digital means, and the need for increased digital autonomy. One example of the former is the conduct of information operations with the aim to influence or destabilise other countries’ political systems. This has been coupled with forced extraterritorial imposition of the laws of certain countries to other jurisdictions. The US, for example, is seeking to set the standard of free speech for the rest of the world. Trump Media and Rumble are prosecuting Brazilian Supreme Court Justice Alexandre de Moraes in a Florida court, challenging his orders to suspend accounts on the two platforms linked to disinformation campaigns in Brazil, a saga being thoroughly analyzed by Curzi. 

These developments relate to the very core of state sovereignty, and should not be mixed up with hedging against digital economic interdependence. They should not depend on the development of ‘sovereign infrastructure’ to be vehemently opposed at present. Countries that are still taking the initial steps in developing their infrastructure have the right to sovereignty, as much as other countries have the duty to collectively ensure that this basic building block of the international society is upheld, offline or online. 

In the upcoming part 2 of this series, the interplay between this landscape and openness will be discussed, by unpacking what autonomy and sovereignty could mean in the context of each Internet layer. Openness and global flows enabled the development of the Internet as we know it.

Digital Sovereignty and the Open Internet: can they co-exist? The Internet Governance Project (IGP) in collaboration with the Quello Center will gather scholars of differing views to debate this question on 4 April 2025, from 10:00 – 11:00 am EST (15:00-16:00 CEST). Registrations are open: https://bit.ly/4gXZY6b

Note: This is an opinion piece written in an individual capacity. It does not necessarily reflect or represent the position of DiploFoundation as an organisation.