Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
As the war in Ukraine progressed, Russian strikes increasingly degraded Ukraine’s terrestrial communications infrastructure, steadily deepening Kyiv’s dependence on SpaceX’s Starlink. The satellite constellation became indispensable for crucial military functions, including command and control, drone operations, and artillery fire coordination, with no readily available European alternative should access be withdrawn.
This reality exemplifies the idea that today, governments rely on critical infrastructure that is privately owned and globally embedded. But these interdependencies can be weaponized, and state-firm relations can become misaligned. Recent controversies over the ability of American and Chinese technology companies to restrict access to critical AI models, alongside NVIDIA’s geopolitical maneuvering, illustrate how privately controlled technologies can restrict state decision-making.
This paradigm is particularly severe in Europe, where external digital infrastructure and semiconductor companies play an increasingly consequential role in the defense, semiconductor, and telecommunications sectors, and by extension, in E.U. geopolitics and national security. The European Union seems to acknowledge its deficiencies and seeks greater control over critical digital networks and technologies. It has even tried to incentivize the private sector to do so through initiatives like the first European Chips Act in 2023. But these have not delivered the necessary support, and so the private sector continues to lament European legislation and regulations.
Currently, there is a poor understanding of the tensions between governments’ strategies that regulate industries for more national resilience and security, and private tech companies that operate more along commercial logic. Without such a clearer account, the political mechanisms underpinning state-business relations in digital infrastructure resilience will remain opaque — even though they are central to its success.
Europe cannot just use financing alone to secure private-public critical infrastructure alignment. It should instead set up mechanisms, including conditionality on public money, procurement used as demand-shaping leverage, and binding rather than voluntary coordination. The European Tech Sovereignty Package, proposed in June 2026, and the anticipated European Savings and Investment Union could provide building blocks for such an approach.
European states currently depend on foreign countries and their private tech firms for more than 80 percent of their critical digital products, services, infrastructure, and intellectual property. Europe lacks sufficient access to semiconductor production capacity and the domestic industrial capacity and resilient supply chains needed to mitigate disruptions in global chip production.
Europe’s pursuit of greater autonomy and sovereignty, and thus greater control over technologies, is not costless. Much of the capability that underwrites European security — from Starlink to advanced lithography — is currently supplied through allies or partners like the United States, Taiwan, Japan, and South Korea, which also underwrite European security. Deliberately reducing reliance on these technologies risks interoperability frictions and higher costs without a guaranteed security return. The case for recalibrating state-firm relations therefore rests on managing exposure at the margin, not on wholesale decoupling from allied suppliers. This requires distinguishing benefits from mutual beneficial interdependence within a technology alliance from dependencies that leave Europe with few alternatives during moments of political or economic crisis.
Recognizing this, the European Commission proposed in June an ambitious Tech Sovereignty Package to develop and control key technologies, infrastructure, and data with less reliance on non-E.U. suppliers. The package also concerns the “front-end” infrastructure that shapes how digital interactions occur, like AI and cloud infrastructure.
The package should, in theory, facilitate governments in mobilizing industry, promoting resilience, and crafting better policy. And yet, European businesses lament government initiatives and subsidies, saying that they create uncertainty, hinder operations, and stifle innovation and competitiveness. In 2025, European and non-European tech leaders pushed back against E.U. regulations for AI systems concerning citizens’ health and safety, leading the Commission to postpone these obligations until December 2027. Europe’s AI ecosystem lost terrain against global competitors, and the E.U.’s rules-setter reputation was harmed.
Businesses pursue their own strategic and commercial goals, which often diverge from governments’ priorities. However, these two crucial actors continue to work at cross-purposes in the effort to improve resilience in digital infrastructure. A European semiconductor company representative once said about the state-business relationship: “[W]e are like players on a football field without a coach that guides us.”
High-end technologies at the frontier of knowledge require special state attention, as they operate in complex ecosystems with many dependencies, very high entry barriers, extreme technological specialization, and businesses with near-monopolistic power. They also attract special security and political attention, making it extremely difficult for states to align policies with these businesses. Effective control over these industries by states in the traditional sovereign sense has moreover become increasingly limited and brings us into a different world of state-business relations.
Europe has some companies — from those that develop highly complex photolithography machines to vacuum valves and world-class research infrastructure — that are chokepoints in global supply chains and can be used as leverage against other states or to shape companies’ behavior like inter-firm trade. But whilst a government may exercise some control over one company, it cannot control the entire network or industry. These industries’ characteristics make it hard, for example, to endogenize the manufacturing capacity of the most advanced nodes, which requires more than financing or political will.
Three sectors best illustrate this difficulty of control: defense, semiconductors, and telecommunications. The first encompasses the overarching structure of modern military capabilities, while the semiconductors and telecommunications represent dual-use “backbone” technologies on which these capabilities rely. Whilst each sector exhibits some similar characteristics, they vary in which characteristic is most pronounced: dependencies in defense, chokepoint power in semiconductors, and infrastructural power in telecommunications.
Defense Sector: Dependencies
In the defense sector, private companies and foreign suppliers own and operate critical technologies and infrastructure, including sensors, drones, cloud, AI, and satellites. These are crucial to the conduct and success of modern military operations, and the European Union heavily relies on foreign suppliers and providers. From the Russo-Ukrainian War to recent conflicts in the Middle East, these capabilities are governed primarily by commercial contracts and corporate decisions rather than the government or military.
Wartime necessities and high attrition rates have driven Ukraine to rapidly scale up its domestic production of commercial tactical drones — it is now manufacturing millions of platforms annually. However, Ukrainian technology companies remain vastly dependent on Chinese suppliers for indispensable components such as lithium batteries, motor magnets, and fiber-optic cabling, with few substitutes at reasonable prices. Kyiv has also been forced to rely on SpaceX’s Starlink satellite constellation to sustain critical military functions. Operational capabilities would be severely degraded were access shut off, as Ukraine lacks European alternatives. Europe’s inability to sustain Ukrainian military aid — lacking the capacity to directly manufacture critical components and ammunitions, or provide key services — puts the continent at risk as it fights to keep Russia from creeping east.
This example underscores three main systemic vulnerabilities and challenges for European countries: a severe scarcity of critical raw materials, protracted industrial manufacturing timelines, and a structural dependence on China and the United States for vital strategic services and military capabilities. These challenges are further compounded by deficits in private investment, fragmented public funding, and weak domestic market demand.
Semiconductor Sector: Chokepoint Power
In the semiconductor industry, no single state — or company — can produce the most advanced microchips alone. Chip design, fabrication, packaging, assembly, and the mining and processing of critical raw materials are distributed across highly specialized firms and multiple countries and continents. This makes semiconductors perhaps the most globally interconnected and relational value chain in the digital economy.
At the center of this ecosystem sits a handful of firms that occupy critical chokepoints. For example, ASML, headquartered in the Netherlands, is the world’s sole producer of the extreme ultraviolet lithography systems required to manufacture the most advanced chips and support the “AI-boom.” Yet, this pivotal position has not translated into strategic autonomy for either the company, the Netherlands, or Europe. The European Union adopted the European Chips Act in 2023, aimed at doubling the E.U. global market share of worldwide production of cutting-edge and sustainable semiconductors from 10 percent to 20 percent by 2030. But the Act was criticized by the private sector, and the Court of Auditor’s special report released in December 2025 stresses that E.U. digital policies and infrastructure are insufficient to achieve the 20 percent target. Today, it still holds around 10 percent market share and lacks the integrated tools needed to achieve more sovereignty in digital infrastructure.
Instead, the E.U.’s central role in the semiconductor value chain has even exposed it to growing geopolitical pressure. Washington used export controls and extraterritorial measures to shape which markets European companies such as ASML can serve, while Beijing has responded by restricting exports of critical raw material inputs like gallium, germanium, and rare-earth processing technologies. As a result, although European governments increasingly regard semiconductors as strategic assets, their ability to mobilize domestic firms remains constrained by global interdependence, commercial interests, and geopolitical competition.
Telecommunications Sector (5G): Infrastructural Power
Unlike the fragmented semiconductor sector, the 5G telecommunications value chain is highly hierarchical. A small oligopoly of vertically integrated system integrators dominates the network’s core technologies and standard-setting, leaving states with few alternative suppliers.
From 2016 on, Beijing’s market penetration in Europe and Washington’s subsequent warnings to Europeans about the security risks of relying on Chinese vendors left E.U. countries caught in the crossfire between the two powers. By 2022, Chinese vendors supplied over half of the 5G radio access network equipment in eight E.U. member states.
In response, European states struggle to mobilize the domestic presence of champions like Nokia and Ericsson and enact a cohesive European response. On the one hand, Chinese offerings were simply too competitive, given that Huawei and ZTE delivered comparable technology at lower prices than European or American alternatives thanks to government support and subsidies. On the other hand, European governments’ room for maneuver was limited because Washington demanded a ban on Chinese vendors, and Beijing threatened retaliation if they complied.
The E.U.’s 5G toolbox is one of its flagship initiatives to strengthen digital infrastructure resilience. While the European Commission provided a common strategic framework, implementation remained largely dependent on member states. The framework could not change the fact that Chinese equipment already embedded in national 4G and 5G networks would be too costly and slow to “rip and replace.” The monetary costs would be distributed across operators, national governments, and E.U. funds. By contrast, the political and diplomatic costs would fall almost entirely on the governments banning Chinese vendors, exposing them to potential retaliation from Beijing. Moreover, the initiative remained voluntary, limiting incentives to coordinate. The situation has not improved — in 2026, China threatened the European Union with serious consequences after Brussels imposed stricter cybersecurity regulations.
Europe’s central challenge is therefore to ensure that public support for more strategic autonomy and technological sovereignty actually translates into greater alignment between private capabilities and public security objectives.
Subsidies are, to an extent, indispensable given the high upfront investments and long development cycles in these high-tech industries. But subsidies are politically unsustainable. While the European Chips Act 2.0 outlines financing instruments, its budget remains uncertain amid the Multiannual Financial Framework negotiations for 2028–2034, and parliamentary and Council discussions. Moreover, the European Chips Act 2.0 is expected to be adopted only in late 2027, while 2021–2027 funding is ending and regional budgets are largely exhausted.
Europe should also mobilize private capital. The anticipated Savings and Investment Union could unlock financing from pension savings, venture capital, and private equity. This would mean Europe has a unified capital market akin to America, improving returns on citizens’ savings and unlocking risk capital for European companies. But financing is insufficient, because it does little to address the underlying problem: Critical capability still sits with industry that states can influence but not fully control. Europe should therefore pair financing with three mechanisms that bear more directly on alignment and control.
First, public money should come with strategic conditionality that public funding secures more than just expanded semiconductor capacity, and instead fosters resilient supply chains and grants governments greater leverage over strategically critical capabilities. The European Chips Act 2.0 takes first steps in this direction. For example, its chapter three allows government procurement procedures to ask applicants to submit a security of supply declaration to strengthen the European semiconductor supply chain. However, this remains voluntary. Firms receiving E.U. support under the European Chips Act 2.0 should be required to demonstrate that additional European production capacity reduces dependencies on single suppliers or countries elsewhere in the value chain. It would ensure that public funding secures more than just expanded semiconductor capacity, fostering resilient supply chains and granting governments greater leverage over strategically critical capabilities.
Second, governments should use procurement as demand-shaping leverage. For example, Ukraine depends on Starlink and Chinese drone components. Still, European investors will remain reluctant to finance development and production facilities of such capabilities where long-term demand is uncertain. The proposed European Tech Sovereignty Package seeks to translate earlier achievements — like the 2023 European Chips successes in mobilizing capital and creating semiconductor pilot lines — into an economically viable technology ecosystem, shifting emphasis from supply-side interventions towards demand-side measures. For example, it links semiconductor policy to AI factories, data centers, cloud infrastructure, and broader industrial applications.
However, whilst the initiative has already been broadly welcomed by both public and private stakeholders, it remains uncertain how much of the proposal will survive consultations and negotiations with the European Parliament and the Council before its expected entry into force in 2027. In response, procurement rules for E.U.-funded or jointly procured defense capabilities could therefore require diversification away from single-country sourcing and toward multi-country sourcing for critical components. A joint European procurement vehicle could simultaneously guarantee predictable, long-term demand for strategically important capabilities and reduce market uncertainty for investors, while requiring participating manufacturers to demonstrate credible alternative supply arrangements. By signaling sustained public demand, governments can improve the commercial viability of alternative technologies and production capacity while giving governments greater leverage over availability and interoperability.
Third, Europe needs enhanced coordination where fragmented national decisions create collective vulnerabilities in the telecommunications sector. The 5G toolbox illustrates the limits of voluntary coordination: Governments agreed on the risk but retained different incentives to remove high-risk vendors. For critical telecommunications infrastructure, common European requirements on vendor diversification and high-risk suppliers should replace voluntary implementation.
European policy should purchase more than just additional capacity. Measurable resilience, credible alternatives, and greater alignment between corporate decisions and European security objectives should be policy priorities. Europe’s dependence on privately controlled and globally embedded technologies cannot be resolved through subsidies and wholesale decoupling alone. Across defense, semiconductors, and telecommunications, the central challenge is to align commercial incentives with public security objectives while preserving benefits of international collaboration: Strategic conditionality can ensure that public funding reduces critical dependencies; joint procurement can create predictable demand for alternative capabilities; and binding European coordination can prevent fragmented national decisions from generating collective vulnerabilities. Together, these mechanisms would give governments greater leverage without requiring direct control over firms or industries, turning technological sovereignty from aspiration into a more credible and sustainable European strategy.
Michiel Foulon, Ph.D., is a guest professor of International Politics at the University of Antwerp and a senior associate fellow at the Royal Institute for International Relations in Brussels. His research interests include European and international security, the geopolitics of technology, and state-business relations, with a particular focus on digital infrastructure, technological sovereignty, and geopolitical competition. His work has been published in leading journals, including Contemporary Security Policy, European Journal of International Security, International Studies Review, and European Journal of International Relations. He is also the founder of Navirix, an ETH Zurich spin-off.
Ivan Zaccagnini, Ph.D., is a senior researcher in Technology and Geopolitics at the Center for Security Studies at ETH Zurich. He is also a non-resident fellow at the Center for International and Strategic Studies at LUISS Guido Carli University in Rome. His research interests include emerging and disruptive technologies, the future of warfare, and great power competition. His work has been published in prestigious journals including International Security, Security Studies, Contemporary Security Policy, Journal of European Integration, and Geopolitics.
Image: Midjourney
