Europe’s Digital Dependence: Can the EU Achieve Strategic Autonomy?

Digital technologies have become increasingly central to economic competitiveness, national security and the functioning of public institutions. As artificial intelligence (AI), cloud computing and semiconductors become embedded in public services, industry and critical infrastructure, access to these technologies is no longer solely an economic concern but also a strategic one. For the European Union, this creates a particular challenge. Europe possesses significant technological capabilities, yet remains dependent on non-European actors for several critical components of the digital economy.

These dependencies are particularly visible in AI, cloud computing and semiconductors, which are closely interconnected. The development of advanced AI systems depends on access to powerful computing infrastructure and specialized chips; cloud services rely on sophisticated hardware and semiconductor technologies; and all three are increasingly essential to Europe’s wider digital ecosystem. The concentration of critical capabilities among a limited number of external providers can therefore expose European governments and businesses to supply-chain disruptions, geopolitical tensions and potential technological or commercial pressure.

This raises an important question: to what extent can reducing the EU’s dependence on non-European technological providers strengthen European strategic autonomy• This article argues that reducing technological dependence can significantly strengthen the EU’s strategic autonomy by increasing resilience, improving control over critical infrastructure and strengthening Europe’s bargaining power. However, strategic autonomy should not be understood as technological self-sufficiency. Given the global and highly interconnected nature of technological supply chains, complete independence would be neither realistic nor necessarily desirable.

Instead, the EU’s objective should be to reduce critical vulnerabilities, diversify external partnerships and develop sufficient domestic capabilities and credible alternatives to prevent excessive dependence on any individual external actor. European strategic autonomy should therefore be understood not as the elimination of technological interdependence, but as the capacity to remain resilient and retain meaningful choices within it.

Mapping the EU’s Technological Dependencies

The EU’s technological dependence is particularly visible in artificial intelligence, cloud computing and semiconductors, although these vulnerabilities are closely interconnected. In artificial intelligence, Europe possesses significant research capabilities and emerging companies, yet remains behind the United States in the development and commercialization of frontier models. In 2024, US-based institutions produced 40 notable AI models, compared with three in Europe[1]. This illustrates Europe’s weaker position in the development and large-scale production of frontier AI models. This gap in frontier-model production is compounded by dependence on the computing infrastructure required to develop advanced AI systems. Training such systems requires specialized GPUs produced by a small number of non-European companies; NVIDIA controlled around 95 percent of the data-center AI accelerator market in 2023[2]. AI dependence is therefore not limited to models themselves, but also extends to the hardware and computing infrastructure on which advanced AI development relies.

Cloud computing represents a second, closely related vulnerability. In 2025, 52.7 percent of EU enterprises used paid cloud computing services, making cloud infrastructure increasingly important to the European economy and to many AI applications[3]. In 2024, Amazon Web Services, Microsoft Azure and Google Cloud together accounted for around 70 percent of Europe’s infrastructure- and platform-as-a-service market[4]. This concentration creates risks of infrastructural dependence, particularly because the European Commission has identified entrenched user bases, lock-in effects and high switching costs among leading cloud providers[5]. Dependence can therefore persist even where alternative services exist if European organizations face significant technical, financial or contractual barriers to switching providers. The issue is consequently not simply the presence of foreign providers in the European market, but the extent to which European users retain meaningful choice between them.

Semiconductors reveal a different but complementary vulnerability. Their production is distributed across a highly fragmented global value chain, and Europe lacks sufficient capacity in some of its most advanced stages. According to a 2025 European Parliament analysis, Europe accounts for around 10 percent of global semiconductor production, compared with 54 percent for Taiwan and 16 percent for China. Europe also lacks leading-edge fabrication capacity for semiconductors below 10 nanometers[6]. At the same time, Europe possesses strategically important capabilities within the semiconductor value chain. Dutch company ASML is currently the world’s only supplier of EUV lithography systems, which are used to manufacture leading-edge chips[7]. This illustrates the specific nature of European technological dependence: the EU is neither technologically self-sufficient nor without strategic capabilities, but occupies critical positions within value chains that remain heavily dependent on external actors for other essential stages.

Taken together, these cases demonstrate that European technological dependence does not result simply from a lack of technological capabilities. Rather, it arises from the uneven distribution and concentration of critical technologies, infrastructure and production capacity across global value chains. AI, cloud computing and semiconductors form an increasingly interconnected ecosystem in which vulnerabilities in one area can reinforce those in another: advanced AI depends on computing infrastructure and specialized chips, while cloud services depend on the hardware and technologies that underpin digital infrastructure. The challenge for European strategic autonomy is therefore not to eliminate technological interdependence, but to reduce those dependencies that could leave European actors excessively exposed to external disruption or pressure.

The EU Response: Building Capacity, Alternatives and Regulatory Power

The EU has responded to these vulnerabilities through a combination of industrial policy, infrastructure investment and regulation. Rather than seeking to eliminate foreign technology from European markets, its approach increasingly aims to strengthen European capabilities, reduce excessive dependencies and ensure that European actors retain meaningful alternatives. This approach was brought together more explicitly in the European Technological Sovereignty Package presented by the European Commission in June 2026, which links initiatives in semiconductors, artificial intelligence, cloud computing and open-source technologies within a broader strategy to strengthen Europe’s digital autonomy and resilience[8].

A. Building European Technological Capacity

The first response has been to strengthen Europe’s capacity to develop and produce critical technologies. In semiconductors, the Commission’s proposed Chips Act 2.0 seeks to reinforce the European semiconductor ecosystem, expand advanced chip production and reduce strategic dependencies and supply-chain vulnerabilities[9]. Importantly, the proposal is not limited to increasing production within Europe. It also seeks to reinforce existing European strengths and maintain Europe’s position within global semiconductor value chains while improving resilience.

A similar approach is visible in artificial intelligence. Through the EuroHPC network and AI Factories, the EU is expanding access to high-performance computing infrastructure for European researchers, start-ups and industry[10]. The Commission’s InvestAI initiative, launched in February 2025, aims to mobilize €200 billion for AI investment, including a €20 billion facility for AI Gigafactories[11]. In July 2026, the EU subsequently expanded its plans to establish up to seven AI Gigafactories, backed by up to €10 billion in European and national funding[12]. These facilities are designed to provide computing capacity at a scale suitable for the development and training of advanced AI models in Europe.

Such initiatives can reduce vulnerability by giving European companies and researchers greater access to critical computing and production capacity. However, investment alone does not guarantee technological autonomy. AI Gigafactories will not by themselves eliminate Europe’s dependence on globally sourced processors and other technologies. The effectiveness of European capacity-building will therefore depend on whether it creates durable capabilities and reduces critical dependencies rather than simply relocating parts of an otherwise foreign-dependent technological ecosystem.

B. Creating Alternatives and Reducing Dependence

The second approach is to reduce the extent to which European organizations are locked into individual technology providers. In cloud computing, this does not necessarily require the EU to create a European equivalent of every US hyperscale. Instead, European policy increasingly focuses on interoperability, portability and the ability to switch between providers. This is important because technological dependence can persist even when alternative services exist if switching remains technically, financially or contractually difficult.

The European Commission’s recent actions on cloud computing illustrate this shift. In November 2025, the Commission launched three market investigations into cloud computing under the Digital Markets Act, including investigations into whether Amazon Web Services and Microsoft Azure should be designated as gatekeepers[13]. In June 2026, the Commission announced its preliminary view that AWS and Microsoft Azure should be designated as gatekeepers, highlighting their entrenched user bases, lock-in effects and high switching costs[14].

The objective is therefore not simply to replace foreign providers with European ones. It is to create a more competitive and diversified environment in which European organizations can change providers and avoid becoming excessively dependent on any single company. This represents a broader conception of strategic autonomy: maintaining meaningful technological choice can itself be a source of autonomy, even when European actors continue to rely on foreign technologies.

C. Regulation as a Tool of Strategic Autonomy

Finally, the EU uses its regulatory power to shape the conditions under which foreign technologies operate within the European market. The AI Act establishes a common regulatory framework for the development and use of artificial intelligence, while the Data Act includes provisions designed to facilitate access to data and switch between cloud providers. The Digital Markets Act, meanwhile, provides the Commission with tools to address market power and practices that can restrict competition and contestability[15]. The Commission’s cloud investigations demonstrate how these regulatory instruments are increasingly being applied to technologies that are strategically important to Europe’s digital infrastructure.

The European Technological Sovereignty Package presented by the Commission in June 2026 brings this regulatory approach together with industrial policy. Alongside the proposed Chips Act 2.0 and Cloud and AI Development Act, the package includes an EU Open-Source Strategy intended to reduce dependencies across the technology stack[16]. This broader conception of sovereignty shows that Europe does not need to control every technology itself to strengthen its strategic autonomy. It can also improve its position by shaping markets, reducing lock-in, promoting interoperability and ensuring that European users retain access to credible alternatives.

Taken together, these policies suggest that the EU is pursuing strategic autonomy through capacity-building, diversification and regulatory leverage, rather than technological isolation. The approach recognizes that Europe cannot realistically reproduce every element of the global technology ecosystem domestically. Its objective is instead to strengthen critical capabilities, reduce excessive dependence on individual external providers and preserve the ability to choose between different technological partners. The effectiveness of this strategy therefore depends not simply on the scale of European investment or regulation, but on whether these measures create genuine alternatives and greater resilience in practice. This question is central to assessing the extent to which reducing technological dependence can actually strengthen European strategic autonomy.

To What Extent Can Reducing Technological Dependence Strengthen Strategic Autonomy?

Reducing technological dependence can significantly strengthen European strategic autonomy, but its effects are best understood in terms of resilience, capacity and freedom of choice rather than complete independence. The cases of AI, cloud computing and semiconductors suggest that reducing critical dependencies can improve Europe’s ability to withstand external pressure and retain control over strategically important technologies. However, these measures cannot remove Europe from the global technological networks on which its economy continues to depend. The extent to which dependency reduction strengthens autonomy therefore depends on whether it creates genuine alternatives and greater room for European decision-making.

The first benefit is greater resilience. Dependence on a limited number of external suppliers can expose Europe to geopolitical tensions, export restrictions and supply-chain disruptions. This is particularly relevant to semiconductors, where production capacity and specialized technologies are concentrated among a limited number of countries and firms. Strengthening European capabilities while diversifying external suppliers can reduce the impact of disruptions affecting any individual source. Resilience, however, does not require Europe to produce everything domestically. A diversified network of suppliers can provide greater protection while preserving access to technologies and expertise that Europe cannot efficiently reproduce itself. The objective is therefore to reduce the consequences of dependence rather than to eliminate interdependence altogether.

Second, greater European capacity can increase control over critical infrastructure and data. Expanding European computing infrastructure, semiconductor capabilities and cloud alternatives can give governments and businesses greater control over how essential digital services are operated and where sensitive data are processed. This is particularly relevant to cloud computing, where dependence involves not only market concentration but also legal, jurisdictional and security considerations. The Commission’s Cloud Sovereignty Framework reflects this broader conception of sovereignty by assessing cloud services across dimensions including legal and jurisdictional exposure, data and AI, supply chains, technology and security[17]. Strategic autonomy therefore involves more than European ownership of infrastructure: it requires European actors to retain sufficient legal, operational and technological control over critical systems to avoid excessive dependence on decisions made outside the EU.

Third, reducing dependency can strengthen European bargaining power. Strategic autonomy is partly a question of having credible alternatives. If European governments and companies can choose between multiple providers, maintain diversified supply chains and rely on domestic capabilities for some critical technologies, external suppliers have less ability to impose unfavorable conditions or exploit European vulnerabilities. This does not require replacing every American or Asian provider with a European equivalent. Rather, the existence of alternatives can itself change the balance of power between European users and external suppliers. In this sense, autonomy can result from having the capacity to say “no” to a particular provider without losing access to an essential technology.

However, the extent of these gains is constrained by several structural limitations. Complete technological independence is unrealistic. Semiconductor production, AI development and cloud infrastructure rely on highly specialized and internationally distributed supply chains. Attempting to reproduce every stage within Europe would require substantial financial resources and could undermine the benefits of international specialization. A realistic conception of technological sovereignty must therefore remain compatible with international partnerships and continued participation in global markets. The objective is not to eliminate external dependencies altogether, but to ensure that those dependencies do not become sources of excessive strategic vulnerability.

Europe also faces a significant scale disadvantage. US and Asian technology companies benefit from larger markets, substantial private investment, established technological ecosystems and economies of scale. European initiatives can strengthen domestic capabilities without necessarily closing the broader competitive gap. This is particularly relevant in AI, where access to computing infrastructure is necessary but not sufficient to produce globally competitive frontier models. Building European computing capacity may therefore improve resilience and bargaining power even if Europe continues to lag behind the United States and China in technological scale and commercialization. Strategic autonomy and technological leadership are consequently not synonymous: Europe can become more autonomous without becoming the world’s leading technological power.

Finally, technological dependence can shift rather than disappear. Increasing European production does not automatically create autonomy if that production remains dependent on foreign inputs. A European AI ecosystem may still rely on foreign processors and GPUs; European cloud infrastructure may depend on non-European hardware; and European semiconductor production may require specialized equipment and components originating elsewhere. Europe’s position in semiconductors illustrates this particularly clearly: it lacks leading-edge fabrication capacity while simultaneously occupying a strategically important position through ASML’s lithography technology[18]. Technological autonomy must therefore be assessed across entire value chains rather than by looking only at where final products are manufactured. The proposed Chips Act 2.0 reflects this logic by seeking not only to increase European production but also to strengthen the resilience of the wider semiconductor ecosystem and reduce critical dependencies[19].

The extent to which reducing technological dependence strengthens strategic autonomy is therefore substantial but conditional. It can increase resilience, improve control over critical infrastructure and data, and strengthen Europe’s bargaining position, even without eliminating its reliance on foreign technologies. The key distinction is between reducing vulnerability and achieving self-sufficiency. European strategic autonomy will be strongest when the EU combines sufficient domestic capabilities with diversified international partnerships and credible alternatives, ensuring that no individual external actor can exercise excessive leverage over Europe’s technological ecosystem. In this sense, dependency reduction rather than dependency elimination provides the most realistic path towards greater European strategic autonomy.

Conclusion

Reducing the EU’s technological dependence on non-European providers can significantly strengthen European strategic autonomy, but not by achieving technological self-sufficiency. AI, cloud computing and semiconductors reveal interconnected vulnerabilities that Europe is addressing through investment, infrastructure, diversification and regulation. These measures can increase resilience, strengthen control over critical infrastructure and improve Europe’s bargaining power by providing credible alternatives to external providers.

However, complete technological independence remains unrealistic given the global and interconnected nature of digital supply chains. European initiatives must therefore focus on reducing critical vulnerabilities rather than eliminating all external dependencies. Strategic autonomy requires sufficient capabilities in strategically important areas, diversified international partnerships and the ability to switch between providers when necessary.

The goal is therefore not to become independent of the global technology ecosystem, but to remain part of it without becoming strategically dependent on it."
References: