Governments across the world are weighing up their exposure to foreign technology platforms, out of concern for data security, supply chain fragility, and fraught international relationships. In Europe, this has produced a “digital sovereignty” doctrine that aims to produce a competitive technology sector.
The European Commission’s June 2025 definition frames digital sovereignty as the “ability to build capacity, resilience and security by reducing strategic dependencies, preventing reliance on foreign actors and single service providers, and safeguarding critical technologies and infrastructure.” The Berlin Declaration, issued in November 2025, sharpened this further, describing sovereignty as the “ability to act autonomously and to freely choose their own solutions.” Both definitions are sound in principle but translating them into policy will not be without risks.
A shared concern, different responses
Europe is not alone in grappling with these questions. Canada’s Digital Charter, updated throughout 2024, reflects similar anxieties about dependence on US-headquartered platforms for critical public services. Ottawa has tended to favour data localisation requirements and procurement preferences, rather than the layered certification architecture that Brussels appears intent on pursuing. The Canadian approach has attracted criticism from trade partners who see data location rules as equivalent to a non-tariff barrier.
India’s Digital Personal Data Protection Act, which came into force in 2023, takes a more selective line, allowing cross-border data flows by default, unless there are specific government restrictions. New Delhi has been explicit that it wants regulation to enable a domestic data economy, rather than constrain one, and the policy design reflects that priority. Singapore’s Digital Economy Framework, widely regarded as a model for smaller open economies, takes a different path again, emphasising interoperability and mutual recognition agreements.
Who bears the compliance burden?
European businesses operating in the technology sector are already navigating the AI Act, GDPR, the NIS2 directive, the Digital Operational Resilience Act, and a series of national cloud certification frameworks. Individually, each of these laws addresses a reasonable concern; together they impose a significant compliance burden. Ironically, the large incumbents, such as the US hyperscalers whose dominance the EU is trying to reduce, are those that can absorb these requirements, while start up and small companies frequently cannot. The Draghi report on European competitiveness, published in 2024, was clear on this point, noting that the regulatory environment has become a source of competitive disadvantage relative to the United States and, increasingly, to parts of Asia.
Infrastructure: the constraint that regulation cannot fix
The European Union currently accounts for roughly only 5% of global AI compute capacity, compared with approximately 75% concentrated in the United States. This reflects complex and slow permitting processes, grid congestion and structural energy cost disadvantages that make Europe a more expensive location for power-intensive computing workloads. Dublin and Amsterdam, two of Europe’s principal data centre hubs, have both had to pause approvals for new facilities because the local grid cannot support additional demand.
The Special Compute Zones proposed in the EU’s AI Continent Action Plan are, however, a step towards righting this wrong as they constitute pre-permitted sites with fast-track grid connections designed to reduce the time between investment decision and operational capacity. Australia has adopted similar arrangements for renewable energy zones, with a single-window permitting model that has materially reduced development timelines.
Conditions for sovereignty
European venture funding grew only 9% year-on-year in 2025, while in North America, venture investment grew 46% over the same period. The gap is not new, but it has widened as AI has become the dominant investment theme and the scale of individual rounds has expanded. For a founder building an AI company in Europe, the difference between a domestic funding environment and a North American can determine whether a business reaches the scale at which it becomes genuinely competitive internationally.
Europe’s predominantly pay-as-you-go public pension systems generate relatively little investable capital compared with the funded models in North America, where pension capital has historically been a significant source of venture and private equity investment. This is not a problem that regulation alone can address, but gradually reforming pension systems and actively incentivising institutional investors to allocate to domestic venture capital is within the scope of policy and has been recommended repeatedly in competitiveness reviews. In addition to this, the history of government-championed technology in Europe, from the telecommunications sector in the 1980s to more recent national cloud initiatives, suggests that insulation from competition tends to reduce the pressure to innovate rather than create healthy competition.
What sovereignty actually requires
Europe has real assets: universities and research institutions that are among the world’s leaders, advanced connectivity infrastructure and technology companies of genuine global scale in enterprise software, semiconductors, and financial technology. Faster permitting for infrastructure, cheaper and cleaner energy, deeper capital markets, and a regulatory outlook that distinguishes between rules that stimulate growth and rules that add friction are what is missing.
The only version of digital sovereignty worth pursuing is one that creates the conditions for European companies to build top-level technology and compete in open markets, domestically and internationally. That means accepting competition, investment, and openness. The choice for European policymakers is between a form of sovereignty that protects incumbents and one that builds challengers.
Daniele Viappiani, Economics and Venture Capital Adviser, GC1 Ventures
