EUROPEAN COMMISSION

Brussels, 17.6.2026

COM(2026) 288 final

COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS

State of the Digital Decade 2026: Closing structural gaps and mobilising investments for 2030 and beyond

{SWD(2026) 154 final} - {SWD(2026) 155 final} - {SWD(2026) 156 final} - {SWD(2026) 157 final}

1.Introduction

In 2026, the EU is facing a combination of longstanding and emerging challenges to its economic and societal prosperity, competitiveness and strategic independence. Rapid technological change, intensifying global competition, geopolitical tensions and rising security risks are increasing the strategic importance of digital policy. Digital transformation is no longer only a matter of innovation and productivity; it is also increasingly tied to resilience, security and democratic stability in line with EU values. Meeting these challenges will largely depend on the success of a whole-of-society digital transformation, underpinning competitiveness, resilience and security while reducing excessive dependencies and strengthening the EU’s technological sovereignty. This requires action across the full digital chain: from education, foundational research and innovation to industrial scale-up, infrastructure deployment, adoption across the economy, effective public sector digitalisation and development of the skills and safeguards needed to ensure that digital transformation works for people.

The 2026 State of the Digital Decade Communication goes beyond stocktaking and identifies structural factors at the root of EU’s digital performance, priority reforms and investments. It concludes that, while tangible progress has been achieved since 2022, it remains insufficient to meet the EU’s objectives and address the above-mentioned challenges. The EU has advanced in a number of areas, including basic connectivity, business digitalisation and the deployment of common digital infrastructures. However, significant gaps persist in foundational technologies, computing capacity, cybersecurity, advanced digital uptake, digital skills and scale-up capacity, as well as inclusiveness and accessibility of digital technologies. These gaps point to four main aspects necessary to increase the EU’s digital leadership: consolidating the strategic technology base; strengthening the security and resilience of digital ecosystems; accelerating the diffusion and adoption of digital technologies across society and the economy; and ensuring that digital transformation works for people by increasing democratic resilience, digital skills levels and social inclusion.

A central message of this report is that delivery depends as much on investment and implementation as on strategy. Closing existing structural gaps will require a change in both the scale and coordination of investment and reforms. The coming period will therefore require a more coherent framework linking reforms, national delivery, EU funding instruments and private investment. This, in turn, will require, the EU, industry, Member States, industry and investors to join forces.

Since its adoption in 2022, the Digital Decade Policy Programme (DDPP) 1  has emerged as a strategic anchor for identifying gaps and priorities and as an operational framework for delivering on Europe’s digital ambitions through a robust governance mechanism that drives recommendations, initiatives and coordinated action across the digital policy landscape (Figure 1). The DDPP combines common objectives and targets with a continuous monitoring cycle, Member States’ involvement, notably through National Digital Decade Strategic Roadmaps (national roadmaps) and EU-level coordination with Member States through the Digital Decade Board. The proposed framework for the next Multiannual Financial Framework (MFF) 2  has further recognised the role of the Digital Decade in supporting Europe’s sustainable prosperity, competitiveness and technological sovereignty by providing that measures financed through the National and Regional Partnership Plans (NRPPs) should contribute to the implementation of recommendations issued under the Digital Decade Policy Programme, while digital investments supported by the forthcoming European Competitiveness Fund (ECF) should address the gaps and priorities identified in the State of the Digital Decade reports.

Figure 1: The Digital Decade objectives and targets and how they translate into EU policies and actions

Further detailed analyses are presented in the annexes and in the supporting documents of this Communication, which together constitute the 2026 State of the Digital Decade Report. This Communication is accompanied by Annex 1 with horizontal recommendations and Annex 2 including the country-specific recommendations, which present a focused and evidence-based set of substantiated priorities for action for the next programming period 3 . Together, they are intended to support a more coordinated, sustained and operational digital policy response capable of strengthening Europe’s competitiveness, technological sovereignty, resilience and capacity to act.

Finally, this report also lays the groundwork for the review of the DDPP later in 2026, with possible proposals for changes to targets, objectives and governance in light of recent technological and EU policy developments.

2.Tracking the overall progress of the EU’s Digital Decade

The strategic context: technology rivalry, AI acceleration and the competitiveness imperative

In 2025, three developments raised the strategic stakes of the EU’s digital transformation.

First, intensifying technological rivalry 4 , the continued consequences of Russia’s war of aggression against Ukraine, the de-risking of relations with China, and renewed transatlantic tensions exposed a core vulnerability: Europe’s dependence on external actors for critical digital infrastructure, technologies, and services. Semiconductor supply chains, cloud capacity, AI ecosystems, and connectivity have become matters of economic and national security for Member States. At the same time, the cybersecurity threat landscape has worsened in both scale and sophistication, with AI amplifying adversarial capabilities and increasing systemic vulnerabilities. Digital technologies have also become key force multipliers in defence, with civilian innovation, notably AI, advanced connectivity, cloud, and autonomous systems increasingly enabling security and defence capabilities.

Second, 2025 marked a further step-change in AI, both in deployment and impact, and the EU is now better equipped than at any point since 2022 to play a leading role in this transformation. AI is shifting from experimentation to widespread adoption across the economy and society, with investment surging to unprecedented levels and AI capturing a very large share of global venture capital in 2025 5 . There has been an historic increase in IT global investments, up 13.5% from 2025 and reaching some USD 6.31 trillion in 2026 6 , notably related to AI with focus on data centres and software (model development). This has led to unprecedented pressure on primary resources, notably chips, storage and high-bandwidth memory. Disruptive advances, including highly autonomous agentic systems capable of tasks such as vulnerability discovery and cyber operations, carry systemic economic, security and strategic implications. Against this backdrop, the EU has substantially strengthened its readiness to compete: the AI Continent action plan, the Apply AI Strategy, the rollout of AI factories and the planned AI Gigafactories, the Cloud and AI Development Act, and the trust framework set by the AI Act together provide a coherent foundation that few other jurisdictions can match in scope. This readiness is matched by societal demand: more than half of EU citizens already use AI tools (59% in their personal lives; more than one third in their work 7 ), while around four out of five Europeans support the development of EU-owned AI tools and believe that AI should be carefully regulated for safety, even if this slows down developers 8 .

Third, the competitiveness and resilience imperative has acquired new urgency. The EU’s labour productivity gap relative to the United States remains close to 20%, driven primarily by gaps in the production and adoption of digital technologies 9 , while structural pressures on energy systems, capital markets and industrial supply chains are reshaping the conditions in which European firms compete globally. Recent assessments by the European Investment Bank (EIB), the Centre for Economic Policy Research (CEPR) and the Annual Single Market and Competitiveness Report converge on the need for a transformative shift in the scale and coordination of investment in digital and complementary technologies. Without such a shift, the EU risks missing the productivity dividend that digital transformation could unlock and deepening its structural dependencies in critical inputs and infrastructures.

What the 2025 metrics reveal about the EU’s digital trajectory

Against this background, the 2025 metrics show a mixed picture. The EU has performed strongly in 2025 in business digitalisation and the deployment of basic connectivity infrastructures. Overall technology adoption (cloud/AI/data) accelerated, with AI recording a 48% increase between 2024 and 2025. The digital intensity of SMEs also continued to improve. Basic 5G coverage has nearly been achieved and fibre to the home (FTTH) deployment has maintained a steady, albeit insufficient, expansion. Moreover, Europe continues to lag behind in terms of take-up of gigabit connectivity, mid-band and 5G standalone (SA). In digital public services, services for both citizens and businesses are projected to reach only around 90% of the target by 2030. Progress in basic digital skills has been strong, bringing the 80% target within close reach by 2037. However, the ICT specialists target has stagnated.

Figure 2: Taking stock of KPIs' progress towards 2030  11   12

At the same time, significant structural gaps persist, notably in human capital, advanced technologies, and scale-up capacity. The shortage of ICT specialists remains severe, as policy efforts made at EU and Member-State level have not yet materialised, and is compounded by persistent gender imbalances. Several weaknesses also persist in the deployment of sovereign critical infrastructures, notably in semiconductors, cloud and AI.

The EU holds only 9% of the global semiconductor market, far from the 20% target, and it continues to underperform globally when it comes to unicorn-sized companies, both in absolute terms and relative to the size of its economy.

Figure 3 shows how the key Digital Decade indicators at EU level have developed between 2021 and 2025, putting the current situation into a four-year perspective. To see current progress towards the 2030 targets, Figure 4 below tracks the progress of KPIs toward the corresponding EU targets, benchmarking achievement levels against projected timelines.

Figure 4: Tracking KPIs progress and expected timeline based on trajectory projections (European Commission’s calculation)

2.3. From data to implementation

Over the past year, cooperation between Member States has translated into more tangible progress, with concrete joint investments and operational initiatives accelerating the deployment of shared digital infrastructures and capabilities across the EU. Through multi-country projects, Member States and the Commission are scaling joint investment in shared digital infrastructures. Two new European Digital Infrastructure Consortia (EDICs) have been set up this year, a further pipeline of new EDICs is advancing, and the Important Projects of Common European Interest (IPCEI) programme is being implemented in microelectronics, cloud and healthcare (including AI-enabled solutions). By operationalising a robust public-private partnership, the AI Gigafactories initiative developed under the EuroHPC Joint Undertaking demonstrates how coordinated action of EU, industry, and Member State efforts can help the EU bridge the gap between digital targets and the necessary scale of implementation. Together, these mechanisms are helping to build strategic capacities that no single Member State could deploy alone.

Progress has also been supported by stronger action and closer collaboration with and between Member States. The Digital Decade Board (DDB) has played a key role in catalysing action. The Board advanced the alignment of the DDPP with the future MFF, prepared the ground for the DDPP review, coordinated updates to national roadmaps, and followed up on the 2025 recommendations. More broadly, the DDB is increasingly functioning as the main national experts’ forum for structured exchange on implementation challenges, coordination needs and simplification.

National roadmaps remain the main vehicle through which this collective effort is translated into national commitments. Through them, Member States are collectively committing to a total of 1 934 measures, representing an overall investment of EUR 289.3 billion, including EUR 205.9 billion from public budgets (equivalent to 1.09% of EU GDP). Member States are also demonstrating their commitment to collective action through the implementation of annual Digital Decade recommendations. In 2025, Member States addressed 64% of the 186 country-specific recommendations issued by the Commission, either through significant policy changes (11%) or making some changes (53%) through new measures. 19 Member States addressed at least half of their recommendations. At EU level, approximately 55% of recommendations show notable (52.2%) or significant progress (2.4%), with particularly strong advances in areas such as uptake of advanced technologies and ICT specialists. However, 45% of recommendations have seen only limited progress.

While momentum is building in several areas, further structured and more ambitious policy efforts remain necessary to accelerate progress. As Annex 1 shows, the gaps in achieving the 2030 targets do not stem from isolated policy shortcomings or short-term cyclical factors. Rather, they point to deeper structural challenges that are closely linked to the EU’s capacity to strengthen its digital leadership. In many cases, these challenges arise in areas where the scale, complexity or cross-border nature of the issues has exceeded the capacity of existing instruments to deliver the necessary coordination and an adequate level of investment. The following section summarises these challenges and their underlying causes. They are also analysed in light of the EU’s strategic priority of establishing Europe as a true AI continent, requiring a comprehensive set of assets, capabilities, and infrastructures, from advanced skills to robust ethical frameworks.

3.Shaping EU’s Digital leadership and technological sovereignty in the AI continent era

In this report, ‘digital leadership’ refers to Europe’s capacity to remain at the technological frontier, shape the development and uptake of digital technologies and infrastructures, and influence the direction of the global digital transformation. It reflects the EU’s ability to leverage the Single Market, innovation, investment and regulation to foster the development, deployment and governance of critical digital capabilities, while promoting international standards and approaches that are aligned with European values and interests.

Achieving this requires coordinated action across several fronts:

·boosting homegrown industrial capacity and autonomy at key steps of the supply chain of digital technologies, progressing towards a full European technology stack while increasing the choice for consumers in the Digital Single Market; 

·securing the supply of digital technologies underpinning Europe’s competitiveness by mitigating dependencies on a single or a limited number of non-EU suppliers;

·gaining control over data infrastructures and critical data while developing the capacity to leverage them effectively;

·setting the standards for key strategic technologies, addressing jurisdictional risks for data and digital service provision, and safeguarding the EU’s capacity to regulate in line with its values and standards; and 

·strengthening the talent pipeline. The structural gaps identified in this section concern precisely those domains where this capacity is most at risk.

3.1Technological sovereignty: closing the R&I and infrastructure gap

The EU still faces the critical challenge of securing control over the full digital innovation chain, from foundational research to large-scale industrial deployment. As part of the Tech Sovereignty Package, the Communication on European Tech Sovereignty  13  outlines the EU’s approach and defines technological sovereignty as Europe’s ability to develop, control and scale the critical technologies, infrastructure, services and data, including digital ecosystems, that underpin its economy, security and society, while de-risking and diversifying supply chains and technological exposure to reduce strategic dependencies and resist foreign interference. 14

3.1.1.Accelerating innovation in digital technologies

European companies continue to lag in digital R&D, particularly in AI. This gap is clearly visible in research and investment trends 15 . EU corporate R&D grew by only 2.9% in 2024, the lowest rate since the pandemic 16 . A persistent sectoral imbalance further disadvantages digital technologies: recent R&D investment increases have been concentrated in energy (+19.8% one year growth rate) and health (+13%), while ICT-related sectors remain comparatively underdeveloped. This situation is exacerbated by the limited number of large ICT firms in Europe, which constrains private R&D, weakens innovation ecosystems and slows the commercialisation of digital technologies.

Global R&D investment is becoming increasingly concentrated among a small number of US technology giants 17  - Amazon, Alphabet, Meta, Microsoft and Apple - particularly in AI, cloud and advanced computing. Over the past decade, these five companies have nearly doubled their share of global R&D spending among the world’s top 2 000 firms, now accounting for around 15% of the total. Without a decisive step-change in investment, coordination and support for scaling European firms, the EU risks further deepening its dependence on external technology providers, with severe consequences for European competitiveness, sovereignty, and security and falling even further behind in the global digital race.

Addressing the scale-up gap requires not only financing instruments but a structural reform of the legal environment in which European digital companies operate. The fragmentation of company law across twenty-seven national jurisdictions creates compliance costs, investor friction and governance complexity that may disadvantage EU-incorporated companies relative to their third country counterparts.

The proposal for a Regulation establishing the 28th regime corporate legal framework (‘EU Inc.’) 18 , adopted by the Commission on 18 March 2026, directly addresses this structural barrier by introducing a harmonised, digital-first and cost-effective set of corporate rules and procedures to facilitate the creation and scaling-up of companies across the single market. A new EU central interface will enable founders and companies to complete incorporation and filing procedures without having to navigate 27 different national systems. The framework will build on existing national business registers and their EU-level interconnection through the Business Registers Interconnection System (BRIS). Realising these benefits will depend not only on the formal adoption of the 28th regime, but also on its consistent operational implementation across Member States, including through clear company formation rules, well-functioning national registries and predictable cross-border enforcement, so that it provides genuine simplification and practical usability for the scale-ups it is intended to serve. 

Open-source software constitutes a further under-exploited sovereignty instrument. Estimated to contribute between EUR 65 and EUR 95 billion to EU GDP and underpinning at least 70% of all code 19 , it reduces dependency on proprietary non-EU systems, lowers barriers for SMEs and start-ups, and can be embedded in public procurement and funding at marginal cost. As its source code is publicly available, it can be freely used, modified, redistributed, and audited. Aligning a greater share of public digital spending with EU open-source development would represent a high-return, low-risk sovereignty dividend. In this regard, the EU Open Source Strategy 20 , published on 3rd June, will review the current state of play and put forward a set of actions to be rolled out in the coming years by the public sector at European and Member-State level, as well as by the private sector and the open-source communities themselves.

3.1.2.Mastering the digital stack: from semiconductors and connectivity to AI infrastructure

Europe’s competitiveness, resilience, and security increasingly depend on its ability to develop and integrate the most advanced layers of the digital technology stack. At the core of this sovereign stack are semiconductors, high-performance computing, cloud and edge computing, AI, quantum technologies, secure connectivity networks, backbone infrastructures, trusted data ecosystems and security-by-design infrastructures. These technologies no longer evolve in isolation. They are tightly integrated and interact closely, shaped by technological progress and the evolving needs of science, industry, public services and defence, including energy efficiency.

In semiconductors, the EU’s share of global value chain revenues stands at about 9%, against a Digital Decade target of 20% to be reached in 2030, in a market where the US, China, South Korea and Taiwan are investing on a scale that risks further entrenching existing asymmetries. This puts the EU in a situation of extreme dependency, as past crises have demonstrated. The Chips Act 21 has so far catalysed thirteen major investment announcements for a total of EUR 32 billion, and, building on its experience, the proposal for a Chips Act 2.0 22  has been adopted to further boost the chips industry, reduce strategic dependencies and support advanced chip production in the EU.

In computing, the EU accounts for only 20% of global data centre capacity, while demand for colocation space is outstripping supply for the third consecutive year, pushing up costs for businesses. The Cloud and AI Development Act proposal 23  aims to triple EU data centre capacity within five to seven years. However, achieving this objective depends on resolving the EU’s persistently higher electricity costs relative to the United States and China, which constitute a structural barrier to data centre investment that the Act alone cannot address.

Quantum technologies are transitioning from research and pilot activities to integration in strategic infrastructures and industrial deployment. The EU is already deploying quantum computers and simulators-integrated with supercomputing centres-through the European High-Performance Computing (EuroHPC) Joint Undertaking. It is also advancing the EuroQCI as a secure quantum communication infrastructure spanning the entire EU, including its overseas territories, and preparing quantum chip pilot lines, design facilities, and quantum internet testbeds. The short-term challenge is to accelerate the move from research excellence to operational capabilities across quantum computing and simulation, communication, sensing, and metrology. This requires support in industrialisation, standards, testing and certification, trusted supply chains, skills development, and demand-side instruments such as public procurement and first-user deployment.

Connectivity remains one of the most critical enabling layers of Europe’s sovereign digital stack and a persistent structural challenge for competitiveness, resilience and security. Despite continued progress in basic 5G and fibre rollout, the EU still lags behind global frontrunners in deploying high-capacity infrastructures, notably full-fibre networks, 5G standalone (SA) and resilient backbone connectivity. The slow transition to 5G SA is a structural constraint, limiting the EU’s capacity to deliver the low-latency, high-reliability services that advanced industrial and AI applications require.

Strategic dependencies in backbone, satellite and submarine cable infrastructures further increase Europe’s exposure in critical data routes, while digital service delivery as a large share of capacity remains under the control of a handful of non-EU operators. Combined with the risk of satellite sabotage or deactivation, this threatens the EU’s economic and security interests. Without accelerated investment, deployment and greater regulatory convergence, these bottlenecks risk undermining the EU’s capacity to scale the next generation of digital, AI and quantum-enabled services on European terms.

The AI dimension adds urgency and acts as the overarching driver of Europe’s digital strategy. While the adoption of AI by businesses has accelerated sharply, reaching 20% in 2025, the EU remains heavily dependent on a handful of non-EU providers for foundational model infrastructure, cloud execution environments and the semiconductor architectures on which AI systems run. Building an AI continent therefore requires simultaneous and tightly coordinated progress across all layers of the digital stack.

The AI Continent action plan 24 provides the strategic framework for this capacity-building effort. In this context, AI Gigafactories - large-scale, shared computing facilities accessible to European researchers, start-ups, and SMEs - are designed to tackle a key bottleneck: the shortage of high-performance compute infrastructure at the scale, the affordability, and security needed for cutting-edge AI development within the EU. Beyond simply providing raw computing power, these facilities are expected to integrate advanced data storage, energy-efficient supercomputing, and secure cloud environments, enabling users to train and deploy large AI models while keeping sensitive data within European jurisdiction. By pooling resources and lowering entry barriers, AI Gigafactories aim to accelerate innovation, support industrial applications and strengthen Europe’s technological leadership in the global AI race. The timely delivery of such ultra-scale sovereign AI compute infrastructure in Europe is a prerequisite for the credibility of the EU’s broader AI strategy, including the sectoral focus of Apply AI Strategy 25 . The EU is also pursuing a comprehensive strategy to strengthen its technological stack: the core objective of the Tech Sovereignty Package 26  adopted on 3 June is meant to strengthen the EU’s capabilities across the entire value chain, moving towards a fully integrated European technology stack.

Five structural factors cut across these technology domains. First, fragmentation: twenty-seven national regulatory and investment frameworks prevent the EU from pooling resources and deploying technology infrastructure at the scale required to compete globally. Spectrum policy, semiconductor R&I funding, and AI compute investment all suffer from the same underlying problem of insufficient coordination. Second, investment asymmetry: in each of these sectors, the EU’s existing instruments, while significant, have not yet been organised around a coherent technology investment logic tied to sovereignty objectives, leaving European support less unified and targeted than what competitor regions have. Third, private-sector underinvestment and more generally the lack of mobilisation of private finance to support the development of the European digital ecosystem, reflecting structural conditions including fragmented capital markets, a weak scale-up ecosystem, and insufficient demand-side incentives for innovation. Fourth, overly concentrated supply chain and dependencies: across chips, network equipment, cloud infrastructure and AI systems, the EU remains dependent on a limited number of external suppliers for components and services that are increasingly treated as strategic assets by competitor nations, and control of them is a lever for geopolitical influence. Fifth, administrative bottlenecks underline the need to accelerate and simplify infrastructure development through facilitated permitting.

Addressing those challenges requires simultaneous and coordinated action across technology policy, investment mobilisation and industrial strategy. Together, the proposals for the Chips Act 2.0, the Cloud and AI Development Act, the Digital Networks Act, and the upcoming Quantum Act proposal will provide an appropriate legislative framework once adopted and implemented. The AI Continent action plan provides the strategic architecture for Europe’s AI capacity-building, with the Gigafactories as its operational centrepiece alongside AI factories, Testing and Experimentation Facilities, and the European Digital Innovation Hubs being repositioned as AI Experience Centres.

Dependency risks also extend to the financial sector. International card schemes account for over 64% of electronically initiated card-based transactions in the euro area 27 , while mobile payments are largely dominated by non-European technology firms, particularly as regards digital wallets. These dependencies create both geopolitical and cybersecurity exposure which could have spillover effects on the financial and economic sector. The digital euro, currently in the interinstitutional legislative process 28 , is a strategic response that will strengthen Europe’s monetary sovereignty and increase the resilience of its payment infrastructure.

Concerns about technological sovereignty also extend to the use of digital technologies in the education sector. European schools and training institutions depend overwhelmingly on non-EU providers for their digital infrastructure - from cloud and operating systems to AI-powered learning tools and learning management systems. The Commission is supporting preparatory work with Member States to set up multi-country projects and collaboration among Member States to strengthen technological sovereignty for schools and training institutions.

Promoting the ethical development and adoption of AI in the cultural and creative sectors is essential to preserving Europe’s cultural sovereignty. AI should be designed and trained in ways that support Europe’s cultural and linguistic diversity, drawing on diverse and representative datasets that reflect the richness of European cultures, while respecting creators’ rights. As announced in the AI Continent action plan, the Apply AI initiative and the Culture Compass 29 , the Commission is preparing a strategy that promotes the development and use of AI in ways that uphold genuine creation, European cultural and linguistic diversity and inclusion.

Legislative and strategic frameworks need to be strengthened by a step-change in investment scale and coordination. The next MFF will play an important role in steering resources towards EU technology priorities, in particular with the proposed European Competitiveness Fund (ECF) and its Digital Leadership window and the new Research and Innovation Framework Programme. The conditions for private R&I at scale must also be improved structurally - through the Savings and Investments Union, the Capital Markets Union, simplified regulatory requirements for R&D-intensive firms, and stronger demand-side incentives including public procurement, as reflected in the Tech Sovereignty Package. The specific investment priorities emerging from this strand are set out in Section 4.

3.1.3 Fostering synergies with defence

The accelerated integration of digital technologies such as AI, advance connectivity or cloud computing into European defence capabilities has emerged as a defining trend of 2025-2026. Technologies originally developed for civilian applications are increasingly dual-use and/or adapted for defence purposes. This ’spin-in’ dynamic means that the EU’s digital industrial capacity and its defence technological capacity are now closely interlinked: both are constrained by weaknesses in semiconductors, computing and cloud infrastructure, AI and secure connectivity.

The EU responded with solid policy and funding architecture in 2025-2026. The Defence Readiness Roadmap 2030, adopted in October 2025, pointed to cyber, AI and electronic warfare as one of nine priority capability areas identified by Member States. The Defence Industry Transformation Roadmap 30  consolidated this direction, framing software-defined warfare, dual-use spin-in and engagement with SMEs and start-ups as cornerstones of Europe’s defence industrial transformation. On the funding side, the European Defence Fund, with a budget of close to EUR 8 billion for the period 2021-2027, is the Commission’s flagship programme in support of defence R&D. As part of it, the EU Defence Innovation Scheme (EUDIS) supports defence innovation and non-traditional defence actors. The AGILE programme, proposed on 25 March 2026 31 , will complement these efforts, with a particular focus on the rapid development and delivery of disruptive defence products and technologies. The European Defence Industry Programme (EDIP) adopted at the end of 2025, delivered its first work programme and launch of calls for proposals at the beginning of 2026. The Security Action for Europe (SAFE), adopted in May 2025, which includes AI and electronic warfare among its eligible procurement categories, has approved national investment plans covering EUR 38 billion of the EUR 150 billion available, with the first disbursements expected in 2026.

Moreover, the mid-term review of cohesion policy 32 (October 2025) enabled cohesion funds to be channelled towards defence and civil preparedness projects. By the same token, the Regulation on incentivising defence-related investments in the EU budget (December 2025) broadened the scope of the Digital Europe Programme to dual-use across all objectives, extended the European Innovation Council (EIC) Accelerator scope to support dual-use innovation, and opened the STEP Scale-Up Scheme to defence. The Commission has proposed that these efforts should continue and be further increased as part of the proposed European Competitiveness Fund (ECF) under the next MFF 33 .

Three structural factors continue to shape the EU’s ability to translate civilian digital investment into defence capability. First, the alignment of strategic priorities between civilian and defence programmes could be further improved across the EU: while EU instruments increasingly recognise dual-use, Member States have yet to systematically embed this convergence in their national roadmaps and national investment planning, and the structural separation between civilian digital and defence innovation communities persists at national level. Second, the industrial base remains concentrated and partly dependent on non-EU providers in semiconductors, cloud and AI. This exposes EU systems to supply disruption and limits the scope for a fully sovereign defence digital supply chain. Third, the security culture and reporting standards across the civilian digital sector are not yet sufficiently developed to make full use of the funding pathways opened by recent regulatory changes. Addressing these challenges requires complementary action and approaches on strategic priorities and on funding-level synergies including across future ECF policy windows, so that investments in computing capacity, connectivity resilience, cybersecurity and AI can consistently support both civilian competitiveness and defence preparedness. Member States are also invited to reflect on dual-use considerations in their updated national roadmaps, and to make use of relevant defence instruments and promote them among their national industrial ecosystems, to participate actively in Capability Coalitions, and to make systematic use of SAFE, EDIP, EUDIS, the dual-use scope of the Digital Europe Programme, the EIC Accelerator and the AGILE programme.

Strengthening the security and resilience of digital ecosystems

The resilience of the digital ecosystem is central to EU’s technological sovereignty as cybersecurity threats becomes more complex, large-scale and systemic. Recent EU data highlight the scale and accelerating nature of cyber risk across the EU: nearly 4 900 significant cybersecurity incidents were recorded in a single year (July 2024-June 2025), pointing to a persistent and evolving threat environment. These attacks are increasingly high-impact, with over 80% of cybercrime incidents involving ransomware, often combined with data theft and multi-layered extortion 34 . Cyber operations are now embedded in state strategies, targeting government, defence and critical infrastructure, while the line between state-sponsored actors and cybercriminals continues to blur through shared tools and infrastructure. Attacks are also becoming more targeted: more than half affect essential entities, while public administrations account for over one third, which means the attacks are increasingly focused on critical sectors such as energy, healthcare, transport and telecommunications, as well as core state functions. At the same time, attacks are becoming more sophisticated and harder to detect, increasingly leveraging AI and combining multiple techniques. Supply chain attacks have emerged as a major systemic risk, whereby a single breach can cascade across multiple organisations due to dependencies on third-party providers and widely used software. This risk is particularly acute for public administrations, which rely on interconnected systems to deliver essential services. Together, these trends illustrate not only the rising frequency of cyber incidents, but also their increasing severity and implications for the EU’s economic security and resilience.

The EU has developed a broad and increasingly robust cybersecurity framework, anchored in NIS2, the Cyber Resilience Act and the proposed revision of the Cybersecurity Act (CSA2) 35 . As part of a wider cybersecurity package, these measures aim to strengthen capabilities and resilience, prevent fragmentation in the digital single market and enhance the security of EU ICT supply chains by ensuring - through a simplified certification framework - that products placed on the market are secure by design. They also seek to facilitate compliance with existing EU rules and increase the role of the EU Agency for Cybersecurity (ENISA) in supporting Member States and the EU in managing cyber threats. In addition, the Cyber Solidarity Act, in force since February 2025, further strengthens collective detection and response capacities at EU level.

However, five structural factors continue to shape the EU’s vulnerability to the evolving cyber threat landscape and limit the effectiveness of its regulatory response. First, implementation fragmentation: while the NIS2 Directive and the Cyber Resilience Act set common requirements, transposition of the NIS2 Directive and enforcement of rules remain uneven across Member States, creating resilience gaps within the single market. Progress also remains uneven across Member States when it comes to restricting high-risk suppliers and preparing for emerging challenges such as post-quantum cryptography. Second, industrial dependency: continued reliance on non-EU vendors (including persistence exposure to high-risk suppliers in critical areas such as 5G networks) exposes the EU to jurisdictional risks and external strategic decisions. Third, a persistent cybersecurity skills deficit, constrains organisations’ ability to operationalise requirements, particularly as AI-enabled threats accelerate. Fourth, AI asymmetry: adversaries are deploying AI faster than it is being integrated into EU defensive systems, creating a structural imbalance not fully addressed by existing frameworks. Fifth, and linked to AI asymmetry, the fundamental shift in cybersecurity brought by the most advanced general-purpose AI models. As these models show unprecedented cyber capabilities, notably in identifying vulnerabilities, and operate at a speed and level of automation that outpaces traditional defensive cycles, we need to stay ahead and ensure our own capacity to use AI as a first line of defence against the possible misuse of such capabilities.

Addressing these gaps requires sustained investment in implementation support, EU-wide interoperability standards, the cybersecurity skills base, and shared detection and response capabilities at EU level. This must be accompanied by full and consistent transposition and enforcement of the NIS2, the Cyber Solidarity Act and the Cyber Resilience Act, the structured phase-out of high-risk suppliers from critical infrastructure and the establishment by Member States of stable, multiannual national funding mechanisms for cybersecurity aligned with EU strategic priorities.

Making digitalisation work for competitiveness: adoption, simplification and decarbonisation

The third structural challenge for digital leadership is to build wider ecosystems and ensure the diffusion of digital technologies across the economy and society. It means ensuring that the digital capabilities being developed in the EU are effectively deployed where they are needed, across businesses, public authorities and society at large, at the speed and scale required by intensifying global competition. Such efforts are critical: so far, the EU has failed to fully capture the gains of the internet revolution, which has led it to lag both in the development of key technologies and in the uptake of digital solutions by public administrations and businesses.

3.3.1. From access to use: accelerating technology diffusion and adoption

Digitally enabled transformation of private and public activities is a catalyst for EU competitiveness. Digitalisation is a key driver of growth and productivity in the EU, and recent evidence points to strong macroeconomic returns on digital investments. The productivity dividend of digital transformation is substantially larger than what the EU is currently capturing, as can be seen in the estimate that 39% of US growth in 2025 was linked to AI technologies 36 . The limited productivity gains from digital investments compared to the US stem not only from differences in ICT capital, but also insufficient investments in intangible assets, such as skills, organisational capacity and management practices, which are essential for the diffusion of digital technologies: between 2009 and 2020, ICT capital accounted for 11.7% of US productivity growth, compared with 3.3% in Europe; over the same period, total factor productivity (TFP), which captures efficiency gains from organisational change and improved resource allocation, contributed 32% in the United States but was negative in Europe 37 .

Broad-based adoption by the general public, enterprises (especially SMEs) and public authorities therefore remains a central challenge. As the KPI data in Section 2 show, AI adoption stands at 20% of enterprises against a 2030 target of 75%, the cloud gap between large firms and SMEs is 36 percentage points, and on current trends the data analytics gap will not close until 2047, while the adoption of gigabit connectivity fast fixed broadband is still low, representing only 27% of subscriptions in the EU in 2025. The main barrier is not only technological availability but access, simplicity and organisational capacity for change. The digital transformation of public services shows similar patterns: while performance at national level is strong (services for the public: 84.6/100; business services: 88.6/100), persistent structural gaps remain in cross-border availability, AI deployment in public administration and the cybersecurity of public digital infrastructure. More than one third of EU government websites are hosted on servers controlled by operators whose ultimate beneficial ownership lies outside EU jurisdiction, and more than half of governmental email domains rely on non-EU operators, creating systemic security exposure and trust concerns that the digital public service agenda alone cannot resolve.

Adoption at scale continues to be constrained by a set of structural factors. A persistent deficit in skills and intermediation limits the capacity of enterprises, in particular SMEs, to identify relevant use cases, ensure compliance and integrate digital technologies into core operations. This is compounded by the limited availability of trusted intermediaries capable of providing integrated technical, legal and business support. At the same time, the pace of organisational transformation remains insufficient. Digital investment does not translate into productivity gains without accompanying changes in business processes, management practices and workforce skills, and evidence indicates that such transformation remains less developed in the EU. Uneven implementation across Member States - in terms of governance, investment levels and delivery capacity - continues to create disparities in adoption.

Public administrations face comparable barriers. While progress at national level is notable, gaps persist in cross-border service provision, the deployment of AI and the security of digital infrastructure. While digitalisation of public services offers significant benefits and opportunities to increase growth through efficiency gains and cost savings for Member States, businesses and individuals, it also requires considerable investment both from the EU and the Member States at the national level to overcome technical difficulties and delays in the delivery of national and cross-border judicial services, as illustrated in the case of digitalisation of justice. Finally, continued reliance on non-EU providers in critical areas creates vulnerabilities affecting both resilience and sovereignty.

Addressing these challenges requires coordinated action across several dimensions. Strengthening skills and intermediation is essential, including through increased investment in digital and AI training and the repositioning of European Digital Innovation Hubs as integrated AI Experience centres linked to AI factories, testing and experimentation facilities and upcoming sandboxes. Greater emphasis must also be placed on organisational transformation, providing the possibility for public support instruments to combine technology uptake with investments in management practices, process redesign and workforce reskilling. Increasing trust requires the development and uptake of secure and sovereign digital infrastructures, including EU-based cloud solutions for public administrations. The EU, Member States, regions and cities should increasingly and systematically leverage public sector demand as a driver of diffusion, notably through strategic public procurement, the deployment of digital public services, and the use of government as a lead customer to create markets, scale solutions and accelerate adoption across the wider economy. Finally, national delivery frameworks should be strengthened through clear governance, dedicated funding and measurable targets, to ensure consistent implementation across Member States, supported by the DDPP.

3.3.2. Simplification: removing the regulatory and infrastructure barriers to adoption

Simplification is itself a critical enabler of technological uptake. Regulatory complexity - particularly the interaction of regulatory obligations, sector-specific requirements and varying national implementations including the introduction of national requirements exceeding those provided for under EU legislation- has been identified by enterprises as one of the primary barriers to digital deployment, especially for SMEs. The Digital Omnibus initiative, which aims to streamline existing data and digital legislation, addresses this directly and is therefore a legislative precondition for accelerating the uptake of digital technologies by enterprises. Moreover, digital, data-driven and AI-enabled solutions for reporting, monitoring and compliance can help reduce this burden, enabling companies to meet regulatory requirements more efficiently while turning compliance into a source of trust, competitiveness and differentiation in global markets.

The European Digital Identity Wallets and European Business Wallets are a central pillar of this strand. By integrating digital verifiable credentials, trusted data sharing, and enabling AI-driven compliance checks, the forthcoming EU Business Wallets will significantly ease how businesses interact with regulators and their customers. Using them will create the conditions for fully digital cross-border interactions - for individuals authenticating to public services and other use cases, for businesses conducting cross-border transactions including e-invoicing, digital contracts and the secure exchange of verified data - and the reduction of administrative burden. All Member States are required to make the Wallet available to private individuals by end-2026. Six large-scale pilots involving over 550 stakeholders across all Member States together with Norway, Iceland, and Ukraine have demonstrated the technical feasibility of cross-border use cases in multiple sectors. Yet deployment remains uneven and the pace of national implementation risks undermining the Wallet’s potential: the certification schemes based on the Cybersecurity Act have not yet been finalised, blocking national deployment timelines; the obligations of the parties relying on the schemes - particularly in the private sector - are still not sufficiently enforced, limiting incentives for the public to adopt the schemes. Accelerating deployment requires the following parallel actions: finalising the certification scheme without further delay; and enforcing relying party obligations systematically, including through the Digital Omnibus where necessary. 

Simplification is also much needed to accelerate connectivity which remains structurally insufficient to support the AI Continent agenda. Regulatory fragmentation with twenty-seven distinct national frameworks for spectrum assignment, network authorisation and different paces of legacy networks migration creates inconsistent investment conditions, prevents operators from achieving the economies of scale required for advanced network deployment and generates legal uncertainty that delays both private investment and public co-financing decisions. The continued operation of legacy copper networks in the majority of Member States negatively impacts demand for full-fibre services, reducing investment incentives, and delaying the migration of users and services to the infrastructure required for next-generation digital applications. Therefore the absence of binding national copper switch-off timelines in most Member States has become a structural rather than a transitional problem. In addition, the absence of long-term and predictable conditions for investments - in particular on spectrum licence duration and renewal - prevents operators from committing capital at the scale and over the time horizons required for 5G SA and 6G deployment. It also limits the ability of public instruments to leverage private co-investment effectively.

For enterprises, especially SMEs, the main barriers to AI and advanced digital adoption are linked to the difficulty of translating interest into deployment safely, lawfully and at an acceptable cost. Fragmented and poor-quality data, weak interoperability with legacy systems, lack of in-house skills, the complexities of aligning early compliance investments with the shifting implementation details of the AI Act, and the absence of trusted intermediaries capable of providing integrated technical, legal and business transformation support remain the proximate constraints behind the SME adoption gap. The operational priorities are fivefold.

·First, the European Digital Innovation Hubs, repositioned as AI Experience Centres and linked to AI factories and Testing and Experimentation Facilities, need to reach their full potential in providing SMEs with integrated technical, legal and business transformation support.

·Second, the deployment of common European data spaces, linked to AI factories and data labs, must be accelerated: access to high-quality, interoperable data is a prerequisite for SMEs to innovate and scale on a more equal footing with large platforms.

·Third, the Digital Omnibus must move rapidly from proposal to implementation, with Member States ensuring consistent national transposition and the Commission monitoring simplification outcomes against measurable benchmarks.

·Fourth, sustained investment in cross-border digital public service infrastructure - including EU Wallet deployment, the Once-Only Technical System, interoperability frameworks and digital identity building blocks, digital verifiable credentials, data exchange mechanisms, trust services - is required to close the gap between national and cross-border service availability, reducing administrative burden for the public and businesses across the single market. 

·Fifth, the Digital Networks Act, currently under negotiation, should preserve its ambition on indefinite spectrum licences, harmonised authorisation conditions and an investment-friendly regulatory architecture as a structural response to the regulatory fragmentation identified above.

At the same time, the copper switch-off, in other words the planned EU-wide migration from legacy copper to full-fibre infrastructure, must be accelerated across all Member States. Targeted public investment must also complement regulatory reform by addressing market failures in underserved areas, particularly rural and remote regions, where private deployment alone will not deliver the coverage required by 2030.

3.3.3. Responsible digitalisation for competitiveness, resilience and decarbonisation

Digital technologies - including AI, digital twins, smart grids and sensor networks - can deliver decisive gains in industrial productivity, energy efficiency, grid optimisation and resilience and circular business models, lowering input costs for EU firms, easing pressure on imported energy and critical raw materials, while accelerating the transition to a more autonomous and resilient industrial base. In a sector as strategic as energy, preserving the Union’s capacity to develop, control and operate these critical digital technologies and infrastructures is an important component of Europe’s technological sovereignty, economic security and long-term resilience.

These positive contributions need to overcome the major resource constraints associated with the expansion of digital infrastructure. Digital infrastructure is increasingly affecting water availability, land use and energy systems at regional level. By 2030, data centre electricity consumption in Europe is expected to rise towards 115 TWh, an increase by at least 45 TWh compared to 2025, driven primarily by AI demand 38 . The EU currently recovers less than 1% of rare earth elements from end-of-life digital equipment and remains highly dependent on external suppliers for several critical raw materials used in digital technologies.

These developments are taking place against a backdrop of tightening energy constraints. The mainstreaming of AI is driving a rapid rise in electricity demand for computing power and data infrastructure 39 . With rising geopolitical tensions, the EU also faces persistently higher electricity costs than the United States, which benefits from lower-cost domestic energy sources, and China, which relies on state-supported supply. Without swift and sufficient expansion of affordable, low-carbon energy and grid capacity, these energy constraints risk undermining the EU’s ability to deploy advanced AI systems at the scale required to maintain competitiveness in the global digital economy.

The EU has set a policy framework to address the digital-green nexus. The proposed clean digitalisation targets - covering data centre energy consumption, net ICT emissions and the contribution of digital solutions to greenhouse gas reductions - provide a necessary basis for making this nexus measurable and operational. The Strategic Roadmap for Digitalisation and AI in the energy sector 40 plans measures to prepare for the increased complexity of the future energy system, including both challenges and opportunities linked to the integration of data centres in the EU energy system and the large-scale AI deployment across the energy value chain. Complementary instruments - including the Cloud and AI Development Act, the Digital Product Passport 41 and the EU Codes of Conduct on Data Centres 42 and on telecommunications sustainability 43  - further strengthen this framework. However, while the policy framework is largely in place, its translation into coherent investment signals, governance mechanisms and measurement systems remains incomplete.

This gap reflects four structural factors limiting the alignment between digital transformation and decarbonisation. First, a governance lag: digital technologies, especially AI, evolve faster than regulatory and investment cycles, delaying the management of their environmental impacts. Second, a growing tension between compute capacity and energy sustainability, as expanding data centre infrastructure increases electricity demand and requires closer alignment between digital, environment and energy policies, including in rural areas. Third, persistent dependencies on critical raw materials, combined with limited recycling capacity, create structural vulnerabilities not yet fully integrated into investment strategies. Fourth, insufficient measurement frameworks: the lack of harmonised EU methodologies to assess both the footprint of digital infrastructure and its contribution to emissions reduction limits effective investment and progress tracking. Addressing these structural factors requires the adoption of new clean digitalisation targets, the development of harmonised metrics, and stronger coordination between digital, energy and environmental policies, with the aims of enhancing EU competitiveness and reducing dependencies on energy and critical materials.

3.4. Ensuring digital transformation works for people

Ensuring that the digital transformation works for people is a core objective of the Digital Decade. The benefits of digital technologies should be accessible to all, while preventing the emergence of new forms of exclusion and protecting citizens and democratic societies from online harms, including disinformation, manipulation and other forms of abuse of digital technologies.

In this respect, digital skills are both a sovereignty asset and an adoption enabler, and the EU’s shortfall in both dimensions, basic skills and ICT specialists, is acute.

Despite significant national investment commitments, the EU remains off-track to meet the 2030 target for basic digital skills (80% of the population), which is currently projected to be reached in 2037. Basic digital skills stand at 60% in 2025 and are forecast to reach only 68% by 2030. The shortfall is concentrated among older individuals, those with low educational attainment, and people in rural areas, precisely the groups for whom digital exclusion carries the highest economic and social costs. A first structural challenge relates to the limited reach of existing provision systems. Current delivery mechanisms do not adequately reach the groups most at risk of lacking basic digital skills - including older people, persons with disabilities, marginalised groups facing discrimination, individuals with low educational attainment, those in rural areas who may face additional challenges linked to limited broadband access. These are largely centred on formal education and workforce-based training, while those most lacking in digital competence are also the least connected to the institutions capable of providing support at scale. A second challenge concerns the effectiveness and level of ambition of current investments. Member States have collectively committed around EUR 24 billion and 349 measures to basic digital skills. However, progress remains below the pace required to meet the target, indicating that while the current distribution of investment, predominantly focused on formal education and general digital inclusion, remains essential, it does not sufficiently address the specific barriers faced by hard-to-reach groups. A third challenge arises from the rapid development of digital technologies. The current target was defined against a pre-generative AI baseline, while the widespread use of such technologies, around 64% of young people aged 16-24 in 2025, is redefining the skills needed for effective and safe participation in the digital environment, indicating the increasing need for AI literacy. This creates a risk that even achieving the target may not create the level of competence required in practice by 2030.

Under the Union of Skills and the Council Recommendation on Human Capital 44 , Member States are encouraged to strengthen basic skills in education and expand training and lifelong learning. The action plan on basic skills recognises digital skills as crucial in modern society and everyday life, and includes them in the basic skills, alongside literacy, mathematics, science and citizenship. The STEM Education Strategic Plan highlights the importance of increasing the talent pipeline in STEM subjects and proposes a 5% enrolment target in doctoral programmes in ICT by 2030 - including a sub-target for female enrolments. The Commission will propose measures to support digital skills development and the digital readiness of schools in the context of its education package. It will also develop guidance on digital skills assessment to improve the provision of skills in school education. The AI Skills Academy supports the workforce dimension of the AI adoption agenda, while the AI literacy framework for primary and secondary education developed in cooperation with the OECD outlines the main skills young people need to develop at school in order to benefit from the AI transformation in a critical and responsible manner. The Digital Competence Framework was updated in November 2025, particularly to take account of competences required in the face of new emerging technologies such as AI and integrate AI competence across all areas of digital competence.

The ICT specialist gap is even more pronounced, with 10.5 million employed in 2025, just 52% of the 2030 target of 20 million in a context where skills requirements are evolving rapidly, reflecting the pace of technological change and the accelerating diffusion of digital technologies across the economy. The proportion of women among ICT specialists has increased only marginally compared to 2015. Closing this gap requires a response commensurate with its scale: including dedicated investments, as well as national interim targets in the updated national roadmaps.

A number of structural factors continue to constrain the EU’s capacity to expand its ICT workforce at the pace required. First, the supply of new entrants remains insufficient. In 2023, the EU produced only 2.7 tertiary ICT graduates per 1 000 young people, compared to 3.7 in the United States and 4.6 in the United Kingdom. This gap cannot be compensated for through upskilling and reskilling alone and points to the need for sustained, long-term investment in initial education systems at a scale not yet reflected in current national commitments. Second, the rapid development of technological requirements is creating persistent mismatches between labour market needs and available skills. The growing importance of areas such as AI, cloud security and data engineering is outpacing the capacity of education and training systems to adapt curricula and programmes. This is further compounded by the increasing role of vendor-specific certifications, which can limit skills portability and reduce overall labour market flexibility. Third, progress in addressing the gender imbalance in ICT professions remains limited. This reflects structural barriers across the entire education and career pathway, from participation in STEM education to entry into and retention within ICT careers, indicating that more comprehensive and coordinated measures are required. Finally, demand for advanced digital skills is expanding rapidly beyond the ICT sector itself. The diffusion of digital technologies, in particular AI, across sectors such as healthcare, manufacturing, logistics and public administration is generating additional demand that exceeds current workforce projections. Addressing this challenge requires a broader, economy-wide approach to skills development, going beyond sector-specific policies. It also requires making the EU a global magnet for ICT talent. The launch of the European Legal Gateway Office pilot in India - in February 2026 - is a first concrete step in that direction. Furthermore, the launch of the STEM Education Strategic Plan in March 2025 lays the basis for anchoring STEM as a priority in the EU education and skills policies and provides a policy and funding framework for advancing the quality of STEM education and participation in these sectors.

Ensuring that digital transformation works for people also requires protecting them in the online environment and preserving democratic resilience. Exposure to disinformation, hostile content and online harm is increasing across all age groups: In 2025, 2 out of 3 young Europeans reported exposure to untrue or doubtful content online 45 . This trend affects democratic processes and public health, with generative AI amplifying their scale and sophistication. It also brings the responsibility of digital actors - particularly AI providers and online platforms - to the forefront, raising issues of accountability and liability. 46  There is increasing recognition of the role of design choices-such as algorithms, decision systems, default settings, and user interaction mechanisms-in shaping societal outcomes. While often framed as technical or commercial decisions, these choices warrant regulatory attention to ensure safer and more trustworthy digital environments. The EU's response is anchored in the enforcement of the Digital Services Act (DSA), with the first DSA fine issued in December 2025. It is complemented by broader policy initiatives, notably the European Democracy Shield, adopted on 12 November 2025, and the Cyberbullying Action Plan adopted in February 2026. These frameworks provide an important basis for addressing online risks and strengthening democratic resilience. The priority now is to ensure their effective implementation, consistent application where relevant, and adequate resourcing, in particular for the European Democracy Shield, while continuing to assess whether additional measures may be needed to address emerging AI-enabled manipulation techniques. 

4.Bridging the funding and reforms gap

The structural challenges identified in the previous sections are reflected in the horizontal and country-specific recommendations addressed to Member States in Annexes 1 and 2 to accelerate collective progress toward Digital Decade objectives and targets. These recommendations converge around a limited set of priority areas intended to shape the EU’s digital agenda including through the next MFF, up until 2034.

Priorities include the accelerated deployment of strategic digital infrastructures: AI compute capacity, full-fibre coverage, 5G SA networks, backbone infrastructures such as submarine cables and resilient semiconductor value chains; the reduction of the advanced technologies adoption gap in enterprises, particularly for SMEs and in key sectors such as manufacturing, healthcare and public administration; the strengthening of digital skills, both in basic literacy and in the development of ICT specialists; the completion of the Digital Single Market’s interoperability architecture, including the deployment of the European Digital Identity Wallet and the Once-Only Technical System; and the reinforcement of cybersecurity across critical sectors and public administrations.

Those structural challenges require a strengthened and coordinated policy response, combining reforms and investments at EU, national and regional level.

Public digital funding is a powerful driver of economic growth, productivity and resilience across the EU

With around EUR 133 billion allocated to digital measures in the Recovery and Resilience Plans (RRPs) as of early 2026, Member States have significantly scaled up investments in digital infrastructures, skills, public services and business transformation. These investments are not only supporting progress towards the Digital Decade targets but are estimated to generate substantial macroeconomic gains across the EU economy by 2030

By 2030, digital Recovery and Resilience Facility (RRF) investments are expected to generate a cumulative economic impact of EUR 219 billion within the EU, rising to around EUR 302 billion at global level 47 . This corresponds to a multiplier of 1.5 within the EU and 2.0 globally, significantly higher than the average impact of RRF spending overall. These strong returns reflect the concentration of digital investments in high-productivity sectors and their capacity to enhance innovation, efficiency and competitiveness. In particular, investments in digital skills and in the digitalisation of public services show the highest multiplier effects, underlining their critical role in enabling the effective uptake and diffusion of digital technologies across the economy. 

The economic benefits of digital investment extend well beyond national borders, highlighting the importance of coordinated action at EU level. Of the total EU impact, around EUR 168 billion stems from direct domestic effects, while a further EUR 51 billion arises from cross-border spillovers through trade and value chain linkages. In several Member States, these spillovers significantly amplify national gains, in some cases doubling or tripling the impact of domestic investments. This demonstrates that the single market remains a key transmission channel for digital investment, reinforcing economic convergence and collective resilience.

Digital RRF investments also generate broad-based gains across all sectors of the economy. While the largest impacts are observed in manufacturing, ICT services, professional services, trade and construction, the benefits extend to sectors that do not receive direct funding, through supply-chain linkages, increased demand and productivity improvements. This confirms the role of digital technologies as a general-purpose driver of growth, enabling efficiency gains and innovation across the entire economic system. These estimates capture the effects of digital investments only. Accompanying digital reforms, including in administrative simplification, network deployment and labour market modernisation, are not modelled, although they are likely to strengthen the effectiveness and long-term impact of RRF digital investments. The figures should therefore be understood as estimates of the investment channel rather than of the full macroeconomic impact of the digital dimensions of the RRF. 

The evidence points to four conditions for maximising the impact of digital investment. First, prioritising high-impact areas such as digital skills, advanced digital infrastructures and the digitalisation of public services is essential to unlock productivity gains. Second, investments in infrastructure must be accompanied by measures supporting adoption and diffusion, in particular among businesses and public administrations, so that technological capacity translates into economic value. Third, coordinated planning and implementation across Member States enhances spillover effects and strengthens the overall impact of investment at EU level. Finally, effective governance and timely implementation of funding programmes, building on the experience of the RRF’s performance-based framework, are critical to delivering results.

Importantly, a substantial proportion of the economic benefits generated by digital investments under the RRF accrues to non-EU economies, reflecting both the EU’s openness and its continued dependencies and integration in key segments of global digital value chains. Estimates indicate that around EUR 83 billion of the total impact leaks outside the EU, with China capturing approximately 27% and the United States around 16% of these spillovers. This reflects the EU’s reliance on external suppliers for critical components and services, including semiconductors, cloud infrastructure and advanced ICT equipment. While openness remains a strength of the EU economy, these patterns underline the need to better align the demand generated by public investment with the development of domestic industrial and technological capacities. Strengthening this articulation is essential to ensure that a greater proportion of public funding translates into value creation within the EU, thereby improving Europe’s technological sovereignty, competitiveness and resilience.

Scaling up reforms and investments to accelerate delivery

At the midpoint of its Digital Decade, this report - together with a growing body of recent analysis from the EIB, CEPR, and independent research institutions 48 - confirms that the EU’s digital objectives will not be achieved without a transformative shift in its investment landscape. The scale of the challenge is well-documented and the figures referenced below draw on a range of recent estimates produced by different institutions, with varying scopes, methodologies and time horizons. While they are not directly additive or fully comparable, they point in the same direction: meeting the EU’s digital objectives will require significantly more investment, stronger coordination and improved mobilisation of both public and private capital.

The Draghi report estimated the additional annual investment needed for digital technologies at around EUR 150 billion. More recent assessments, based on a wider set of assumptions-including geopolitical developments and defence spending, suggest that overall EU strategic investment needs could be substantially higher, potentially approaching EUR 1 200 billion annually through 2031. 49  The size of investment required to achieve EU’s ambitions in areas critical to EU technological sovereignty points to a clear need for public and private capital to be mobilised to unprecedented levels. The strengthening the EU’s semiconductor ecosystems alone will require EUR 120 billion. Expanding data centre capacity will require around EUR 200 billion by 2036, plus another EUR 100 billion for the full realisation of the Cloud and AI leadership initiatives, as well as the deployment of AI Factories and Gigafactories. Finally, for all measures under the open-source strategy, an estimated EUR 2 billion will need to be mobilised over the next 7 years. For energy, the annual investment gap is estimated at EUR 400 billion. 50  The 2026 Annual Single Market and Competitiveness Report confirms that the EU’s labour productivity gap relative to the United States remains 20%, driven primarily by gaps in the production and adoption of digital technologies, and estimates that achieving the EU’s digitalisation targets would unlock economic gains equivalent to 1.8% of GDP 51 . According to estimates by the EPRS, advancing high-tech digital innovation alone across Europe will require between EUR 212 and EUR 380 billion annually, more than triple the EU’s current yearly total investment in digital technologies and infrastructure 52 .

The digital transition is a core element of the Commission’s investment strategy to boost competitiveness. Between 2021 and 2025, EUR 229 billion of the overall EU budget (including NextGenerationEU) was dedicated to the digital transition, representing almost 14.5% of the total EU budget for that period 53 . A significant share of the public spending comes from the Recovery and Resilience Facility, which dedicates EUR 133.1 billion to digital measures as of April 2026. Of this, EUR 120.4 billion is estimated to contribute directly to achieving Digital Decade targets 54 .

The most acute near-term risk is the emergence of a significant investment shortfall. It is expected that nearly half of the public budget of the measures included in national roadmaps will be phased out by 2026, with 58% by 2027. This potential gap of one to two years between the expiry of RRF-funded measures and the operational deployment of ECF and NRPPs instruments could substantially reduce investment momentum in key digital priorities. The loss of predictable investment signals in ecosystem building and strategic technology development would carry significant costs for Europe's competitiveness and technological leadership. Building on this report, the Commission will help identify the areas most exposed to such investment gaps and seek to develop a bridging framework thereby enabling Member States to maintain implementation momentum.

Mobilising private investment at scale is a central priority. Closing the investment gap will require stronger synergies between public and private resources, co-investment models and innovative financial instruments, building on successful examples such as EU leadership in high-performance computing (HPC) and the rapid scale-up enabled by AI Gigafactories. The scale of the private financing gap is stark: the EU accounts for only 5% of global venture capital fundraising in innovation, compared to 52% for the United States and 40% for China 55 . European AI start-ups raised approximately EUR 11 billion in venture capital in 2024, one-sixth of what their US counterparts raised 56 . For cybersecurity alone, EU venture capital totalled EUR 814 million compared to EUR 15 billion in the United States 57 . Closing this gap requires the deepening of the Savings and Investments Union, the simplification of Initial Public Offering (IPO) pathways, and the deployment of risk-sharing instruments 58 , particularly for deep-tech investment where traditional bank financing remains structurally insufficient and ensuring that both private and public operators can internalise digital sovereignty as a measurable risk dimension.

However, relying solely on market forces will not be sufficient: all major economies are deploying large-scale public funding, including sovereign investment and equity participation in strategic sectors. Japan’s support for semiconductor initiatives in its first development phase illustrates the scale and strategic nature of such interventions 59 . Europe must respond with a funding strategy that is similarly targeted and sufficiently large to be efficient, focused on addressing market failures, crowding in private investment and fostering cooperation with SMEs and start-ups, while preserving sustainable public debt levels 

Strategic public procurement constitutes one of the most direct and under-exploited demand instruments available to Member States. The forthcoming revision of the Public Procurement Directives, combined with GovTech approaches 60 and the proposed Cloud and AI Development Act 61 , will provide the regulatory basis for public administrations to systematically integrate security of supply, strategic dependency and resilience considerations into the procurement of critical digital solutions, notably for cloud infrastructure, AI systems and cybersecurity.

The next MFF (2028-2034) provides a critical opportunity to support reforms and investment in digital policy and accelerate delivery. Structural reforms were already a central and legally required component of the Recovery and Resilience Plan (RPP) framework 62 . The RRF demonstrated the value of combining investment with reform: of approximately 2 500 digital milestones and targets in recovery plans, around 24% qualify as structural reforms 63 , covering network deployment, cybersecurity frameworks, labour market modernisation, capacity building investments and the once-only principle. National and Regional Partnership Plans will help the EU achieve its policy objectives more efficiently by linking EU funding to reforms, thereby enhancing the EU’s leverage to encourage and assist Member States in overcoming institutional and regulatory obstacles that hinder the fulfilment of EU policy priorities. Reforms can also help to increase the positive impact of investments, hence increasing the value of each euro spent. Such Plans will support the digital transformation towards the Digital Decade targets and objectives set out in the Digital Decade Policy Programme 2030. The ECF will also play a central role in supporting the digital transformation of the Union.

As proposed by the Commission, the link between the DDPP and the investment architecture of the next MFF could not only be strategic but also operational. Through its monitoring and gap analysis, the DDPP can help identify, - in complementarity to the European Semester and other relevant policy analyses, such as the post-2027 Common Agricultural Policy recommendations and National Energy and Climate Plans - investment and reform priorities that should be reflected in both EU-level instruments and national plans.

Figure 6: EU-relevant budget for the Digital Decade targets (2021-2027 MFF, incl. RRF 2020-26, EUR million, cut-off date April 2026 for RRF and March 2025 for all the other instruments)  64

The analysis across Digital Decade KPIs and objectives can help inform the targeting of investment, by identifying areas of structural underperformance where EU-level action offers the greatest added value. Moreover, the Digital Decade recommendations will provide Member States with an evidence-based basis for shaping national programming for digital issues - fostering complementarities and synergies between EU and national level efforts. The Digital Decade tools will also contribute to inform the work of the Competitiveness Coordination Tool (CCT), through which the EU, Member States and industry are already joining forces to strengthen strategic EU-wide value chains within common competitiveness priorities. This new way of working, which is currently being used to deploy AI Gigafactories, should maximise impact through coordinated investments and reforms.

National roadmaps are the key instrument for translating these priorities into concrete commitments. They set out how Member States plan to deliver on Digital Decade targets, ensure continuity of investment beyond the Recovery and Resilience Facility, and contribute to both national and EU-level priorities. Following the DDPP rules, Member States are required to update their national roadmaps by December 2026, explicitly linking adjusted and planned measures to Member State-specific recommendations. Member States will also be encouraged to indicate whether, and to what extent, the planned measures are aligned with ECF priorities and could potentially benefit from the future NRPPs to facilitate and prepare the ground for the future MFF. These updates should take due account of the recommendations set out in Annex 2 and contribute to a more coherent, sustained and impactful implementation across the EU.

The local and regional dimension: a governance gap to close

Cities and regions are central to the delivery of the Digital Decade, as they deploy infrastructure, provide digital public services and translate EU policies into tangible outcomes for people in the EU. However, the current governance framework does not fully reflect this role. Local and regional authorities are not systematically involved in national roadmap design/update processes and territorial data remains insufficiently integrated into monitoring. Meanwhile, limited administrative and digital capacity at local level continues to be a key implementation constraint.

This gap has direct implications for delivery. Evidence from the Commission’s November 2025 workshop on the role of cities and regions, as well as from initiatives such as the LORDIMAS framework and national mapping processes involving thousands of municipalities, indicates that the main barriers to local digital transformation lie in governance arrangements and funding structures, rather than in political commitment or innovation potential. At the same time, initiatives such as the CitiVERSE EDIC demonstrate the benefits of structured cooperation when appropriate frameworks are in place.

Moreover, recent analytical work at subnational level confirms that digital divides are increasingly multidimensional and cannot be fully captured by national averages. Regional disparities persist not only in infrastructure but also in capabilities, usage and socio-economic outcomes 65 .

The forthcoming review of the DDPP and the proposal for the next MFF provide a key opportunity to address these shortcomings. The Commission will work to strengthen the role of local and regional authorities in governance processes, improve the integration of territorial data into monitoring, and ensure that EU and national funding instruments adequately support local and regional digital capacity.

5.Conclusion and next steps

The 2026 State of the Digital Decade report marks a turning point in the EU’s digital governance framework, as the analytical work of previous years is now potentially matched by new, stronger operational tools to act on it, notably in the context of the next MFF

However, acting on it effectively requires three commitments from all involved. First, a commitment to scale: the investment levels required to close the structural gaps identified in this report are of an order of magnitude that neither individual Member States nor EU instruments can deliver alone, and that can only be achieved through genuine coordination and co-financing, including both private and public funding. Second, a commitment to speed: the pace of technological change means that instruments designed for a 2022 context must be recalibrated for a 2026 one, and the forthcoming DDPP review is the vehicle for that recalibration. Third, a commitment to coherence: the three structural challenges identified in this report - building strategic technologies, deploying them widely, and ensuring they work for people - must be addressed simultaneously, not sequentially. The stakes of delivering on these commitments are significant: an ambitious and coordinated digital policy response achieving the set ambitions, could raise EU GDP by up to 1.8%, reflecting gains in productivity, more efficient public service delivery, and a strengthened innovation ecosystem 66 . This would require significantly scaled-up EU-level investment capacity, stronger coordination of public investment and RDI, reduced fragmentation, higher investment efficiency and improved mobilisation of public and private funding.

Delivering on these objectives requires renewed collective commitment from Member States, EU institutions and stakeholders at all levels. Particular attention must be paid to the risk of a severe investment shortfall. Since about half of the public budget of the digital measures currently included in national roadmaps is expected to be phased out by the end of 2026, and about six out of ten by the end of 2027, Member States and the Commission must jointly act to avoid a significant public funding gap and preserve momentum in support of EU’s digital leadership, competitiveness and technological sovereignty. The December 2026 revision of the national roadmaps is the political window to address this issue.

The revision of the National Digital Decade Strategic Roadmaps in December 2026 provides the first opportunity for Member States to translate the priorities set out in the SDD2026 package into concrete national commitments and prepare a pipeline of mature projects for implementation under the next Multiannual Financial Framework. The Commission calls therefore on all Member States to use the December 2026 revision of their national roadmaps as the central political vehicle for closing the gaps to their 2030 targets, ahead of the preparation of, and for synergies with, the next MFF. Each revised roadmap should respond to recommendations of Annexes 1 and 2 with concrete measures detailing reforms or investments, timelines, impacts, and budgets while building on existing strategies, fostering synergies with the ECF, and integrating local and regional dimensions 67 . The Commission stands ready to support Member States throughout this process, notably through the Digital Decade Board, and will report in the 2027 State of the Digital Decade on the collective progress achieved based on the revised national roadmaps.

The evidence gathered in this report should also inform the forthcoming review of the DDPP. This will be a key opportunity to adapt the framework to technological developments, streamline its instruments, and strengthen its governance, notably by strengthening the link between strategic priorities, funding mechanisms and implementation tools.

(1)

European Parliament and Council of the European Union, Decision (EU) 2022/2481 of 14 December 2022 Establishing the 2030 Policy Programme "Path to the Digital Decade" ’, OJ L 323, 19 December 2022.

(2)

European Commission, A European budget fit for Europe’s ambitions: Multiannual Financial Framework 2028–2034, legislative package adopted on 16 July 2025.

(3)

The Digital Decade Report issues recommendations under Decision (EU) 2022/2481. Article 3 defines its general objectives, including resilience, reducing the digital divide, technological sovereignty, digital capability deployment, digital empowerment, cybersecurity, and sustainability. Article 4 sets specific targets for Member States in areas such as skills, infrastructure, and the digitalisation of businesses and public services. Article 6 mandates the Commission to assess annual progress toward these objectives and targets, and to recommend actions for Member States. Article 7 lays down national Digital Decade strategic roadmaps as the main tool for implementing these recommendations.

(4)

 OECD, Science, Technology and Innovation Outlook 2025 , 28 October 2025.

(5)

 OECD, Venture Capital Investments in Artificial Intelligence through 2025 , OECD Policy Briefs No 50, February 2026.

(6)

 Gartner, Gartner Forecasts Worldwide IT Spending to Grow 13.5% in 2026, Totalling $6.31 Trillion , 22 April 2026.

(7)

GONZALEZ VAZQUEZ, I., FERNANDEZ MACIAS, E., WRIGHT, S. and VILLANI, D., Digital Monitoring, Algorithmic Management and the Platformisation of Work in Europe, Publications Office of the European Union, Luxembourg, 2025

(8)

 Special Eurobarometer 572]  https://data.europa.eu/doi/10.2759/3200731

(9)

 European Commission, Annual Single Market and Competitiveness Report 2026 , COM(2026) 46 final, 30 January 2026.

(10)

  https://data.europa.eu/doi/10.2759/3200731

.

(11)

The current KPI for the 5G target does not reflect the actual quality of service experienced by users. It monitors areas where a 5G signal is available, regardless of the network performance. Therefore, the current stage of 5G deployment can be considered only as ‘basic 5G’.

Regarding quantum, data is based on: Strategic Advisory Board of the European Quantum Flagship , Key Performance Indicators for Quantum Technologies in Europe , March 2025. By the end of 2025, the EuroHPC Joint Undertaking had procured six quantum computers, two of which were inaugurated in 2025, and two additional analogue quantum simulators were inaugurated under the HPCQS project.

(12)

Full details on the KPIs are in the DESI 2026 Methodological note at  https://digital-strategy.ec.europa.eu/en/news-redirect/938549  

(13)

COM(2026) 503

(14)

  JRC Publications Repository - Open but Not Powerless: Towards a Common Understanding of EU Digital Sovereignty .

(15)

The EU leads in high-impact research in only three of the 74 critical technologies; Australian Strategic Policy Institute (ASPI), Critical Technology Tracker , 2025.

(16)

 European Commission Joint Research Centre, The 2025 EU Industrial R&D Investment Scoreboard , December 2025.

(17)

Ibid.

(18)

Regulation (EU) 2026/74 of the European Parliament and of the Council on the 28th regime corporate legal framework, 'EU INC.'

(19)

Open Forum Europe, cited in: Blind, K. et al., The Impact of Open Source Software and Hardware on Technological Independence, Competitiveness and Innovation in the EU Economy , Final Study Report, European Commission, September 2021. Open-source software is estimated to contribute between EUR 65 and EUR 95 billion to the EU’s GDP and underpins approximately 70% of all code.

(20)

Communication from the Commission to the European Parliament, the Council, the European Economic and Social Committee and the Committee of the Regions on European Tech Sovereignty, accompanied by an EU Open-Source Strategy, COM/2026/503 final

(21)

Regulation (EU) 2023/1781 of the European Parliament and of the Council of 13 September 2023 establishing a framework of measures for strengthening Europe’s semiconductor ecosystem and amending Regulation (EU) 2021/694 (Chips Act)

(22)

COM(2026) 504 final

(23)

COM(2026) 502

(24)

European Commission, AI Continent Action Plan , COM(2025) 165 final, 9 April 2025.

(25)

Encompasses the public sector and 10 key industry sectors: healthcare and pharmaceuticals, mobility, transport and automotive, robotics, manufacturing, engineering and construction, climate and environment, energy, agri-food, defence, security and space, electronic communications and cultural, creative and media sectors

(26)

  The Tech Sovereignty package consists of 2 legislative proposals (the Chips Act 2.0 and the Cloud and AI Development Act, the EU Open-Source Strategy and a Strategic Roadmap for Digitalisation and AI in Energy

(27)

Volume share of international card schemes in total electronically initiated card payments with cards issued in the euro area, first half of 2023. Based on data collected under Regulation (EU) No 1409/2013 of the European Central Bank on payments statistics (ECB/2013/43), as amended.

(28)

 Lane, P.R., ‘ The Digital Euro: Maintaining the Autonomy of the Monetary System ,’ keynote speech at the University College Cork Economics Society Conference, European Central Bank, 20 March 2025; the Commission’s legislative proposal on the digital euro was adopted in June 2023 and continued interinstitutional negotiations in 2024; see: European Commission, Proposal for a Regulation of the European Parliament and of the Council on the Establishment of the Digital Euro , COM(2023) 369 final, June 2023.

(29)

A Culture Compass for Europe COM(2025) 785 final

(30)

 European Commission, EU Defence Industry Transformation Roadmap , COM(2025) 845 final, 19 November 2025.

(31)

European Commission, Proposal for a Regulation Establishing the Programme for Agile and Rapid Defence Innovation (AGILE) , COM(2026) 135 final, 25 March 2026.

(32)

European Commission, A modernised cohesion policy: The mid-term review , COM(2025) 163 final

(33)

European Commission, Proposal for a Regulation Establishing the European Competitiveness Fund (ECF), Including the Specific Programme for Defence Research and Innovation Activities , COM(2025) 555 final, 2025.

(34)

 European Union Agency for Cybersecurity (ENISA), ENISA Threat Landscape 2025 , October 2025.

(35)

 European Commission, Proposal for a Regulation on the European Union Agency for Cybersecurity (ENISA), the European Cybersecurity Certification Framework, and ICT Supply Chain Security, and Repealing Regulation (EU) 2019/881 (Cybersecurity Act 2) , COM(2026) 11 final, 2026.

(36)

 Rubinton, H. and Patro, B.A., ‘ Tracking AI's Contribution to GDP Growth ,’ On the Economy, Federal Reserve Bank of St. Louis, 12 January 2026.

(37)

JRC Report (upcoming).

(38)

 IEA, Table A.4: Data Centres Electricity Consumption by Region, p. 110, Key Questions on Energy and AI , International Energy Agency, April 2026.

(39)

 IEA, Energy Demand from AI , in Energy and AI, International Energy Agency, April 2025. EU data centre electricity consumption is projected to grow from around 70 TWh in 2024 to approximately 115 TWh by 2030.

(40)

COM(2026) 501 final

(41)

Regulation (EU) 2024/1781 of the European Parliament and of the Council of 13 June 2024 establishing a framework for the setting of ecodesign requirements for sustainable products, amending Directive (EU) 2020/1828 and Regulation (EU) 2023/1542 and repealing Directive 2009/125/EC

(42)

https://joint-research-centre.ec.europa.eu/jrc-news-and-updates/eu-code-conduct-data-centres-towards-more-innovative-sustainable-and-secure-data-centre-facilities-2023-09-05_en  

(43)

  https://publications.jrc.ec.europa.eu/repository/handle/JRC144975  

(44)

https://data.consilium.europa.eu/doc/document/ST-6081-2026-REV-1/en/pdf

(45)

 Eurostat, ICT usage in households and by individuals, 2025. Available at: Eurostat, ICT Usage in Households and by Individuals – Evaluating Data, Information and Digital Content, 2025.

(46)

K.G.M. v. Meta Platforms et al., Los Angeles County Superior Court, judgment of 25 March 2026, finding social media platforms liable for harm linked to addictive design features. Liability was based on platform design (addictive features), not just user-generated content

The court held that design features (e.g. infinite scroll, autoplay) were a “substantial factor” in causing harm.

(47)

 Michels, A., Ferreira, V., Annoni, P., Burton, J., Pedauga, L., Rueda-Cantuche, J. M. & Kušen, M., European Economy. Discussion Paper 249:  Digital Measures under the Recovery and Resilience Facility: Economic Impacts at Macro, Sectoral and Country Levels , European Commission, Directorate-General for Economic and Financial Affairs, 2026.

(48)

 See: European Investment Bank, Investment Report 2025/2026: Capitalising on Europe’s Strengths, March 2026; EIB Investment Survey 2025, October 2025; Buti, M. et al., ‘How large is the investment gap in the EU and how to close it?’, CEPR Policy Insight No. 141, 2025; Zettelmeyer, J., ‘Draghi on a shoestring: the European Commission’s Competitiveness Compass’, Bruegel Analysis, March 2026; European Parliamentary Research Service, Cost of Non-Europe in High-Tech Digital Innovation, July 2024; Draghi, M., The Future of European Competitiveness, 2024. For the private investment gap, see McKinsey Global Institute / WEF, Transforming Europe: Bold Moves to Lift a Continent, January 2026.

(49)

 Buti, M. et al., ‘Time to be strategic: how public money could power Europe’s green, digital and defence transitions’, CEPR VoxEU, July 2025.

(50)

COM(2026) 503

(51)

 European Commission, Annual Single Market and Competitiveness Report 2026, COM(2026) 46 final.

(52)

 European Parliamentary Research Service, Cost of Non-Europe in High-Tech Digital Innovation: Investment Needs and Economic Benefits, July 2024, pp. 56-57 and 60-61. This remains the most recent EPRS estimate; an updated analysis is expected in the context of the MFF 2028-2034 debate.

(53)

European Commission, Digital Tracking , accessed 2026.

(54)

Torrecillas, J. (2026). RRF April 2026 Update. JRC146664, European Commission.

(55)

 European Commission, Competitiveness Compass for the EU , COM(2025) 30 final, 29 January 2025.

(56)

Dealroom.co, Opening moves in global AI - AI, startups & venture capital, AI Action Summit, Paris, February 2025.

(57)

2025 SDD report cites: European Commission, European industrial technology roadmap for the next generation cloud-edge offering, 2021 (used for 2024 comparative data). For a more current source, see EIB Investment Report 2024/2025: Innovation, integration and simplification in Europe, 2025, which documents structural cybersecurity financing gaps in comparable terms. Recommend verification before publication.

(58)

European Commission, Savings and Investments Union: A Strategy to Foster Citizens' Wealth and Economic Competitiveness in the EU , COM(2025) 124 final, March 2025.

(59)

 Center for strategic and international studies: Japan Seeks to Revitalize Its Semiconductor Industry , 2023. Also Foreign Policy, Japan Wants Semiconductor ManufacturingBack Home to Frustrate China , 9 January 2023.

(60)

The forthcoming revision of the Public Procurement Directives is listed in the Commission’s work programme for 2025-2026. GovTech refers to public sector engagement with start-ups and SMEs to procure innovative solutions (as defined in the 2025 SDD Communication, footnote 27).

(61)

Cloud and AI Development Act (CADA): Commission proposal to triple EU data centre capacity within 5-7 years, ensuring EU-based, sovereign cloud capacity for critical public sector applications.

(62)

Regulation (EU) 2021/241 establishing the Recovery and Resilience Facility, Article 18(4), requiring recovery and resilience plans to contribute to addressing challenges identified in country-specific recommendations.

(63)

Recovery and Resilience Facility Scoreboard; see also 2025 SDD Communication, section on public and private investment levels.

(64)

The figures include 5 main funding instruments: Recovery and Resilience Facility, Cohesion Policy, Horizon Europe, Digital Europe Programme and Connecting Europe Facility-Digital. Torrecillas, J. (2026). RRF April 2026 Update. JRC146664, European Commission. See also Torrecillas, J. and Nepelski, D., Update of Mapping of EU funds to Digital Decade targets 2021-2027, Publications Office of the European Union, Luxembourg, 2025, https://data.europa.eu/doi/10.2760/4123945, JRC141966.

(65)

 ESPON, DigiReg – Territorial Perspectives of Digital Transition in European Regions , 2024.

(66)

 European Parliamentary Research Service (EPRS), Benefit of an EU Strategic Innovation Agenda – Cost of Non-Europe , EPRS Study PE 762.853, February 2025.

(67)

In line with the Commission’s Roadmap guidance: Guidance to the Member States on the preparation of the national Digital Decade strategic roadmaps C(2023) 4025 final.

EUROPEAN COMMISSION

Brussels, 17.6.2026

COM(2026) 288 final

ANNEX

to the

COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS

State of the Digital Decade 2026 Closing structural gaps and mobilising investments for 2030 and beyond

{SWD(2026) 154 final} - {SWD(2026) 155 final} - {SWD(2026) 156 final} - {SWD(2026) 157 final}

Contents    

1.    Introduction    

2.    Reinforcing technological sovereignty, digital leadership, security and competitiveness    

2.1.    Cutting-edge technological capacities    

2.1.1 Semiconductors    

2.1.2 Quantum    

2.1.3 High-performance computing, AI Factories and AI Gigafactories    

2.1.4 Edge nodes and computing capacity deployment    

2.2.    Supporting innovative companies with digital tools and resilient network    

2.2.1.    Connectivity    

2.2.2.    Digitalisation of SMEs and uptake of digital technologies by EU enterprises: AI, Cloud and data analytics    

2.2.3.    Open Source    

2.2.4.    Unicorns    

2.3.    Ensuring security for competitive growth    

3.    Protecting and empowering people, reducing burdens and harnessing digitalisation for sustainability    

3.1.    Digital skills for smart society and competitive economy    

3.1.1.    Basic digital skills    

3.1.2.    ICT Specialists    

3.1.3.    Protecting people, in particular minors, in the online space    

3.2.    Efficient public services and administrative burden reduction    

3.2.1.    European Digital Identity and business wallets    

3.2.2.    Digital Public Services for Citizens and businesses    

3.2.3.    e-Health    

3.3.    Digital for decarbonisation and sustainable technologies    

3.3.1.    Sustainable digitalisation for competitiveness, resilience and net positive impacts    

3.3.2.    Rising environmental concerns: electricity, water and material demand for digital transition    

3.3.3.    EU actions to unlock the twin green digital transition    

3.3.4.    Member State actions towards the twin green and digital transition    

3.3.5.    Reforms and investments needed to accelerate the green and digital transition    

4.    Funding the Digital Decade    

5.    International    

1.Introduction

This Annex forms an integral part of the State of the Digital Decade 2026 report. It covers in particular technological sovereignty, security and competitiveness; the protection and empowerment of people; the role of digitalisation in the green transition; and a strengthened horizontal dimension on coherence, efficiency and simplification across policies and instruments. It also includes horizontal recommendations.

Recommendations issued under the Digital Decade policy programme may address all dimensions of the Programme: not only the Digital Decade targets, but also the general objectives set out in Article 3 (e.g. sovereignty, resilience, competitiveness, security, fighting digital divides) taking into account the digital principles and rights of the European Declaration. Each recommendation is intended to be operationalised by Member States through the updated National Roadmaps to be submitted by December 2026, in accordance with the coherence framework set out in Section 4 of the Communication.

Member State-specific recommendations complement these horizontal recommendations and are included in Annex 2. The identification of recommendations takes into account the Member States’ performance for these areas (as measured by KPIs or other evidence for areas without KPI) plus a comprehensive policy assessment of the measures taken and/or planned by the Member States for that specific areas informed notably by dedicated bilateral exchanges with administrations, civil society and national regulators taking into account structural factors specific to each Member State, the follow-up to the recommendations issued in 2025 and the measures set out in National Roadmaps with the Programme.

The key objective of the recommendations is to enable collective progress and achievement of DD objectives, requiring two complementary and mutually reinforcing approaches: they address gaps where progress is insufficient, and they leverage strengths identified in a Member State - in line with the cooperative approach of the Programme. Reinforcing and scaling up national strengths can support the European Union as a whole to reach the common Digital Decade targets and objectives (for instance, leadership in technologies which are critical for EU’s sovereignty, e.g. AI, semiconductors or quantum).

The Member State-specific recommendations focus on a limited number of structural priorities - about five per Member States, are designed to be actionable and future-oriented, and are prioritised on a scale according to expected impact and relevance in the MS context. This targeted approach ensures that recommendations concentrate on the areas with the greatest potential impact on the basis of its expected contribution to the collective achievement of the Union's Digital Decade objectives and targets.

The recommendations of the State of the Digital Decade report operate alongside the digital dimension of the Country-Specific Recommendations adopted under the European Semester and other relevant documents, such as the CAP recommendations. These instruments are complementary. On the one hand, the Digital Decade recommendations aim at achieving the EU-level targets and objectives for the digital transformation of the EU On the other hand, the European Semester can identify digital-specific shortcomings with macro-economic, employment or social impact and address recommendations about digital reforms and investment needs in these areas. These elements are also mirrored in the Digital Decade analysis and recommendations, where applicable with additional elements. The Digital decade is also focusing on advanced digital infrastructures deployment linked to EU’s industrial policy and capacities (e.g. semiconductors, quantum) and cybersecurity. The two tracks have been prepared in close coordination to ensure consistency and complementarity.

The 2026 horizontal recommendations also build on the assessment of the implementation of the 2025 EU-level recommendations, carried out under Article 6 of the Digital Decade Policy Programme Decision 1 and presented in the accompanying Staff Working Document. Recommendations assessed as showing limited progress in 2025 are carried over in updated form, while areas showing notable or significant progress give rise to recommendations focused on consolidation, scaling and uptake.

The analysis provides a comprehensive overview of the state of play, identifying key areas of progress and acceleration, as well as persistent gaps, structural weaknesses, and emerging risks. It highlights the Union’s strengths to build on and pinpoints the main bottlenecks requiring reinforced reforms and public and private investment. It also includes updates on the comprehensive monitoring of the Declaration of digital rights and principles undertaken in 2025. The assessment primarily relies on monitoring through the Digital Economy and Society Index (DESI), complemented by relevant studies, expert analysis and the National Digital Decade Strategic Roadmaps submitted by Member States, with a view to informing targeted policy action, improved coordination and stronger collective delivery.

In parallel to investment and capacity-building efforts, the Union has continued to strengthen the regulatory framework underpinning the digital economy, with a view to ensuring fair and contestable markets. In particular, the Digital Markets Act addresses structural imbalances in platform ecosystems by limiting the ability of large digital gatekeepers to act as bottlenecks and by ensuring that digital businesses have opportunities to grow and innovate and together with users can benefit from greater choice, interoperability and access to digital services across the Single Market.

According to the Digital Decade Eurobarometer 2026, three out of four Europeans believe that the digitalisation of public and private services is making their lives easier. Digital health technologies and green digital technologies (e.g. energy-saving tech) are selected by at least half of the respondents when asked which technologies are likely to have a positive impact over the next ten years.

2.Reinforcing technological sovereignty, digital leadership, security and competitiveness

The EU’s ability to develop, deploy, and maintain control over critical digital technologies is increasingly proving fundamental to its long-term competitiveness, technological sovereignty, and strategic resilience. These dimensions are closely interlinked: sustained productivity gains from digital technologies, particularly AI, can only be secured if Europe reduces its strategic dependencies and strengthens the security of its digital ecosystem. In several critical areas, including cloud computing, advanced computing, semiconductors, and high-performance connectivity, the limited availability of competitive European alternatives continues to constrain substitutability.

This increases risks related both to security of supply and to the jurisdiction applicable to data and services when providers or controllers are not established in the EU. Such dependencies weaken the Union’s capacity to regulate and enforce its rules effectively, while exposing key infrastructures and services to external vulnerabilities.

Addressing these risks requires sustained investment in critical technologies, a systematic reduction of strategic dependencies, clear solutions to jurisdiction and enforcement issues for data and services controlled from outside the Union, and robust cybersecurity across the digital value chain - from hardware and infrastructure to applications and services as well as reinforcing international cooperation as appropriate (e.g. to ensure access to markets, secure alternative supply lines, etc). The deployment of critical technologies must be matched by their effective uptake and diffusion across the economy and society, in particular among SMEs, which remain central to unlocking productivity gains and ensuring broad-based benefits from digitalisation. For SMEs and startups to take up opportunities and innovate, rigorous enforcement of the Digital Markets Act is needed to tackle structural imbalances in digital markets, where a limited number of large platforms controls SMEs’ access to end users.

While progress has been made in certain areas, the EU continues to underperform globally in several strategic domains. International competitors are consolidating their own sovereignty and leadership through scale, investment and integrated market dynamics, with global leadership in critical technologies increasingly concentrated in the US and China 2 . Bridging these gaps will require reinforced coordination between the Union and the Member States, together with better alignment of reforms, investments and governance frameworks.

2.1.Cutting-edge technological capacities

Europe’s ability to compete globally rests on its command of the foundational technologies that underpin modern computing: semiconductors, quantum systems, and edge and computing infrastructures. These domains are deeply interlinked: advanced semiconductors drive HPC, quantum and edge systems, edge nodes bring computing capacity closer to where data is generated, and quantum technologies promise to redefine computational limits.

Progress across these three areas remains uneven. The EU’s share of global semiconductor revenues stands at 8.8%, well below the 20% target set for 2030 in the Digital Decade Decision. Quantum has met its Digital Decade milestone, but deployment is held back by fragmentation and by the scale of investment needed to move beyond NISQ systems 3 . Meanwhile, while edge node deployment is on track to meet the 2030 Digital Decade target ahead of schedule, overall computing capacity still lags significantly behind demand and remains well below US levels.

The stakes attached to these KPIs are both economic and strategic: gaps in these areas translate into critical dependencies on non-EU providers, higher costs for businesses and public services, and a diminished capacity to develop and deploy the next generation of digital technologies on European terms.

A defining trend of 2025-2026 has also been the accelerated integration of digital technologies into European defence capabilities. Digital technologies are no longer peripheral to defence: they increasingly act as core force multipliers, reshaping how capabilities are developed, deployed and integrated across domains. Artificial intelligence, advanced connectivity including in space, cloud computing, cybersecurity tools and autonomous systems, originally developed for civilian applications, are now being systematically adapted for defence purposes, strengthening command and control, situational awareness and electronic warfare. This structural shift from civilian to military innovation, often described as "spin-in", has emerged as a key driver of capability development, and is likely to deepen further as operational lessons from recent conflicts continue to inform technology adaptation and procurement priorities across Member States.

2.1.1 Semiconductors

The global semiconductor race has intensified further over the past year. Chips are at the centre of competitiveness, security and resilience strategies worldwide, as they underpin artificial intelligence, cloud and edge computing, future communications networks, software-defined vehicles, industrial automation, medical devices, defence systems and the wider digital and green transitions. The economic and strategic relevance of semiconductors therefore continues to grow, and with it the pressure on governments to secure access to critical technologies and strengthen trusted supply chains. These dynamics are reshaping the semiconductor market with AI emerging as a pervasive end application. AI-related components are expected to drive growth and account for more than 70% of the total semiconductor market by 2030. Over the past two years, market growth has been driven by chips for AI data centres, notably processors and memory, while most other segments have remained stagnant.

The European Union occupies a distinctive position in the global semiconductor ecosystem. Its role is not primarily defined by scale in the most volume-driven segments of leading-edge logic and memory manufacturing, but rather by strong capabilities in strategic parts of the value chain, including equipment, materials, research and technology infrastructures, specialty manufacturing, and semiconductor devices for automotive, industrial, power, sensing and secure applications. This profile reflects the structure of EU industry. The EU hosts leading integrated device manufacturers and specialty foundries, globally relevant equipment and materials suppliers, and major research organisations. As a result, it plays an enabling role well beyond its own regional demand, especially in applications where reliability, energy efficiency, safety, long product lifecycles and system integration are more important than pure scaling at the smallest geometry.

Against this background, EU semiconductor value-chain revenues have increased almost steadily, from EUR 53 billion in 2019 to an estimated EUR 93 billion in 2025, as shown in the figure below which displays the EU's semiconductor value chain market in absolute values and the EU's share from 2019 to 2030. This growth is expected to continue, almost linearly, to EUR 143 billion in 2030.

In 2025 the EU's share of global value chain revenues is estimated at 8.8%, still far from the 20% target to be reached by 2030. This share is projected to remain relatively stable in the coming years, in a context of sustained large-scale investments in other regions of continued expansion in the global market which is now projected to exceed EUR 1.6 trillion in 2030. The EU market continues to grow in absolute terms. Although the EU share values are not increasing, the value of the EU27 semiconductor market has increased by 9.4% since 2024, compared with 8.0% growth in global market revenues (Figure1). EU semiconductor value-chain revenues have also increased by 4.5%, from 89 in 2023 to 93 billion euro in 2025.

Figure 1: EU semiconductors value chain revenues (on the left axis in billion EUR) and global market share (on the right axis in %). The EU market shares are shown in the black labels (Data source: International Data Corporation) 4 .

In their National Roadmaps, in 2025 Member States committed to investing EUR 50.2 billion in semiconductors, accounting for 17% of the roadmaps’ total budget across all targets. Private sources are expected to contribute EUR 8.6 billion to this investment. The 53 measures reported in the roadmaps mainly focus on supporting R&D and on boosting production capacity and industrial deployment of semiconductors. Roughly one-third of the measures are dedicated to each of these areas, pointing to a balanced approach to growth and innovation. These areas also remain a priority in the Member States’ roadmap adjustments.

The European Chips Act 5 has established a framework to attract investments from semiconductor manufacturers into first-of-a-kind EU facilities. Since the launch of the initiative, already 13 projects have been announced representing more than EUR 32 billion in investments, with other promising projects upcoming. The Important Project of Common European Interest on Microelectronics and Communication Technologies (IPCEI ME-CT) is now fully operational. It brings together 14 Member States and 47 companies, channelling around EUR 20 billion from both private and public sources into 57 collaborative projects across several countries, out of the 68 projects initially envisaged (due to some withdrawals or early conclusion). In parallel, the proposed new IPCEI on Advanced Semiconductor Technologies (IPCEI AST) is expected to enter the notification phase shortly.

The Chips Act, through the Chips for Europe Initiative, has also launched a set of initiatives aimed at building technological capacity and accelerating the transfer of innovation from research to industrial deployment. State-of-the-art pilot lines, supported by a total of EUR 3.7 billion funding, offer shared, industrial-scale environments where new technologies can be tested, validated and prepared for production in key areas such as beyond 2nm leading-edge system-on-chip, fully depleted silicon-on-insulator applications, advanced packaging, wide-bandgap materials, and photonic integrated circuits. The design platform focuses on reinforcing the EU’s capabilities in chip design, enabling companies (particularly SMEs and start-ups) to develop more complex and system-level products. Competence centres serve as entry points to expertise, training and technology support, anchoring knowledge in regional ecosystems, and facilitating access to infrastructure and skills across Member States. Competence centres may also support regions within each Member State in developing long-term strategy to host, attract and expand semiconductor-related investments to the benefit of the local ecosystems. Quantum chip pilots prepare Europe for emerging computing and sensing technologies by creating pathways from frontier research to manufacturable components, while the Chips Fund complements these infrastructures by improving access to risk finance and scale-up capital.

For the European Union to compete globally in the semiconductor sector, it is essential that each Member State develops national semiconductor strategy providing strategic objectives, priorities and relevant roadmaps, to substantially increase investments and to continue commitment to the leading value chain areas, including semiconductor equipment, chips design, analogue components, sensors, photonics, while also securing a strong entry into emerging markets such as computing and AI-oriented silicon. With this aim in mind, the Commission has started the formal review of the Chips Act, targeting the Chips Act 2.0 announcement in Q2 2026 with a clear support from industry, and Member States.

Public funding must also strike the right balance between predictability and flexibility, while crowding in the much-needed private investment and avoiding distortions of competition in the internal market. A stable and predictable trajectory is essential for budget planning, but the pace of technological development also requires the capacity to respond swiftly to emerging priorities. In this respect, Joint Undertakings (JUs) are comparatively well equipped, as their procedures allow work programmes to adapt more rapidly when new needs arise. By pooling public and private resources at scale, JUs have played a pivotal role in aligning strategic agendas, and fostering robust ecosystems around key EU policy priorities, thereby strengthening Europe’s competitiveness and technological sovereignty. However, Member States’ financial planning has not always been sufficiently flexible to accommodate emerging needs and changing priorities. In addition, due to the complexity of rules applicable to the blending of funding, Member States’ departments responsible for R&I funding in JUs must ensure close coordination with state aid experts, a process that can lengthen administrative timelines before final decision making.

2.1.2 Quantum

The starting value for this KPI was 0 in 2022 and it reached and surpassed the target in 2024 as the first two quantum simulators were deployed in France and Germany, see trajectory ( Figure 2 ). Additional quantum computers are expected to be deployed before the end of the decade, as several procurements are currently ongoing. Given the specific nature of this target, no baseline trajectory has been established.

In their National Roadmaps, Member States reported investing EUR 4.1 billion in quantum computing (1.4% of the total budget of the National Roadmaps), of which EUR 358 million comes from private sources. The 62 measures reported mainly focus on supporting R&D and the deployment of quantum technologies, with roughly one third of the measures dedicated to each area. In their adjustments, Member States primarily focused on R&D for quantum technologies.

Figure 2: Number of quantum computers in the EU. Trajectory towards 2030.

The target for the Quantum KPI has been reached and the KPI has currently a value of 6, expecting to reach 10 by the end of the decade.

The main challenge for the EU in this area remains fragmentation. Globally, in 2025, it is estimated that there were more than 100 quantum computers 6 . Europe overall hosts around 40% of these systems distributed across more than 20 vendors. However, all these systems are Noisy intermediate-scale quantum (NISQ) systems with a low number of qubits (< 1000) and error rates that do not allow computations demonstrating a clear quantum advantage.

For the EU to lead in this field, Member States need to coordinate their efforts and investments to support the scaling up of the technology. Building a fault-tolerant quantum computer (FTQC), will require a very substantial investment by the EU and the MSs (hundreds of millions EUR 7 ), together with private and venture capital. Such a computer, operating with stable logical qubits, will be capable of running algorithms that demonstrate a quantum advantage.

Progress towards this stage will therefore depend on joint cooperation and sustained investment focused on carefully selected technological priorities. Other countries, notably the United States have launched such initiatives 8  including ones that are also open to foreign companies. If the EU does not act in time, there is a risk that European companies will develop and commercialise their technologies outside the Union or be acquired by foreign competitors, with the related technology and intellectual property subsequently developed outside the EU.

The recently published Quantum Europe Strategy 9  sets out the objectives in this area while the upcoming Quantum Act will provide the legal and budgetary framework for reaching them.

Finally, Europe also needs to increase its share of private investment in quantum technologies. While the EU still leads in terms of the number of investments, in 2025, out of the USD 4.36 billion invested globally in quantum technologies, companies in Europe attracted only 23% of the global amount invested (USD 1.014 billion), compared to 64% attracted by companies based in the United States 10 . Moreover, these private investments in the EU represent only 10% of the total public investments (MS and the EU combined) 11 .

In the domain of quantum communications, the global leader is China with rapidly expanding ground optical quantum networks as well as multiple quantum communication satellites. The EuroQCI 12 initiative aiming at deploying a secure quantum communication infrastructure across Europe is an ambitious European response, encompassed in the IRIS2 Secure Connectivity regulation.

2.1.3 High-performance computing, AI Factories and AI Gigafactories

Since its establishment in 2018, the EuroHPC Joint Undertaking (EuroHPC) has built one of the most powerful infrastructures for high-performance computing (HPC) and artificial Intelligence (AI) worldwide. Together with its participating states, the EuroHPC has acquired nine supercomputers, including three systems-JUPITER (#4), LUMI (#9), and LEONARDO (#10)-ranked among the ten most powerful supercomputers in the world. In September 2025, JUPITER, the first European supercomputer to reach the exascale frontier, was inaugurated. A second exascale supercomputer (Alice Recoque) is to be deployed within 2027. Several additional mid-range EuroHPC systems are currently being installed. These efforts have contributed to the development of a world-leading, secure, and interconnected supercomputing ecosystem, broadening HPC use, and strengthened the skills base for European science and industry.

The EuroHPC Joint Undertaking is rolling out 19 AI Factories and 13 Antennas across Europe, with an overall investment of around EUR 2.6 billion. This involves the procurement of 15 new AI-optimised supercomputers, increasing Europe’s AI computing power fivefold. AI Factories and Antennas will cooperate as a federated network ensuring seamless integration, efficient resource sharing, and secure cross-border access, thereby advancing Europe’s strategic autonomy in critical digital capabilities.

A major scientific success story of 2025 was the advancement of the Destination Earth initiative, which performed frontier high resolution climate simulations primarily using the LUMI supercomputer and its new AI Factory capabilities.

Europe stands at a pivotal moment to convert recent progress in HPC and AI into durable leadership, technological sovereignty, and broad-based impact. Expanding AI computing capacity within the EU remains a top-tier strategic priority to strengthen competitiveness and technological sovereignty. Building on the concept of AI Factories, AI Gigafactories are intended to take this a step further by integrating massive computing power in large-scale facilities designed to develop, train, and deploy the next generation of the most complex AI models at an unprecedented scale.

AI Gigafactories will be selected through an official Call, based on joint procurement between EuroHPC and its participating states. These infrastructures are essential if Europe is to compete at the global level and strengthen its strategic autonomy in science and in critical industrial sectors. Given the scale of investment required, AI Gigafactories are expected to be implemented through public-private partnerships.

A critical factor in the HPC/AI ecosystem remains the strong dependence on third-party sources, notably for HPC and AI chips. For a sovereign ecosystem to thrive, the EU must develop indigenous building blocks (i.e., the necessary hardware and software) to power these supercomputing infrastructures. Europe has strong engineering talent in the field, but producing competitive European alternatives-spanning semiconductor design, packaging, integration, and optimised software stacks-will require sustained funding and industry partnerships well placed to bring these capabilities from prototype to volume deployment.

2.1.4 Edge nodes and computing capacity deployment

Edge computing is emerging as a critical complement to traditional cloud infrastructure. While the current wave of artificial intelligence-particularly generative AI-relies heavily on centralised, high-performance computing resources hosted in hyperscale cloud environments and High-Performance Computing (HPC) centres, an increasing share of data processing is shifting closer to where data is generated. Edge nodes enable low-latency processing, real-time analytics, and more efficient data handling by reducing the need to transmit large volumes of data to distant cloud facilities. In this architecture, cloud and edge operate in tandem: the cloud provides the large-scale computational power required for training and orchestrating AI models, while edge infrastructure supports inference and time-sensitive applications at or near the end user. Beyond incremental optimisation, there is also an evolution towards “native AI,” which requires embedding AI directly into system design rather than layering it onto existing architectures. Although more complex, this approach unlocks significantly greater performance gains-much like 5G standalone delivers far higher capabilities than non-standalone 5G despite sharing the same label.

In 2025 the Edge Observatory methodology has been significantly improved based on the lessons learnt in the previous years of the analysis. The updated methodology is based on a Computer-Assisted Telephone Interviewing - CATI - survey of more than 430 validated respondents across all the EU Member States. The respondents are selected amongst IT infrastructure decision-makers surveyed in their local language. The stratified sample is selected according to geographical distribution, type of industry (according to Eurostat’s official classifications NACE Rev. 2 ), and company size. The platform DataCentreMap 13 is used to monitor public edge nodes. On-premise node figures, limited to enterprises with 250+ employees, are derived on the basis of Eurostat’s enterprise statistics combined with site-to-node ratios by the Germany's Federal Statistical Office, used as a proxy in the absence of equivalent EU-wide data. A series of validation interviews with both CATI respondents and external experts are carried out to cross-check the results, supported by multi-source triangulation and outlier disqualification to ensure statistical robustness. For these reasons, 2025 data cannot be compared with previous years (break-in-series) 14 .

Figure 3: Edge node estimated deployment (EU projection to 2030).

As of end-2025, the Edge Observatory for the Digital Decade estimates that 7 451 climate-neutral and highly secure edge nodes 15 are deployed across EU, spanning four categories: on-premise, far edge, near edge, and in-country edge data centres. Private on-premise near edge (295 nodes), and far edge (67 nodes). 

Despite representing only 6% of nodes in 2025, in-country data centres account for the large majority of total edge capacity - potentially up to 92% of aggregate MW if nodes operate at maximum rated capacity. On-premise nodes, by contrast, despite constituting 89% of all nodes, represent a comparatively modest share of total capacity, given the small footprint of each individual deployment (approximately half a rack).

When sustainability and security criteria are set aside, the total on-premise edge population alone exceeds 17,000 nodes, indicating a large share of deployments that do not yet meet climate-neutral and security standards.

Edge node deployment across the EU27 remains heavily concentrated in larger, high-GDP economies, with Germany leading both on-premise and public edge deployments. In 2025, Germany leads with 1 771 carbon-neutral and secure on-premise edge nodes, followed by France (737), Italy (629), Poland (536), and Spain (529). For public edge nodes, Germany, France, and the Netherlands together account for 51% of all public edge nodes in the EU27, with regional hotspots in West-Nederland (Amsterdam), Hessen and Nordrhein-Westfalen (Frankfurt area), and Île-de-France (Paris), all areas with established colocation and hyperscale data centre infrastructure. Deployments strongly favour metropolitan and industrial zones, while rural areas and transport hubs remain the lowest priorities. Future expansion is expected to broaden geographic and service zone coverage, although targeted rural deployment remains the least anticipated growth dimension.

The total number of climate-neutral and highly secure edge nodes deployed across EU27 Member States is projected to reach approximately 14 000 by 2030, representing an 88% increase relative to the estimated 7 451 nodes deployed at end-2025 16 . On this basis, the Digital Decade target of 10 000 nodes is projected to be reached in 2028, two years ahead of the 2030 deadline.

It is expected that on-premise nodes will retain structural dominance throughout the period, although their relative share is forecasted to decline modestly from 89% to approximately 84% as near edge infrastructure scales more rapidly. Public edge infrastructure (in-country data centres, near edge, and far edge combined) is projected to more than triple from around 1,032 nodes in 2025 to over 3 500 nodes by 2030. This is expected to translate into a significant expansion of aggregate compute capacity at the edge by 2030.

The expansion of edge nodes must be analysed and contextualised within the broader context of cloud infrastructure, which continues to underpin the overall availability of computing capacity. The rapid advancement of AI is fuelling an unprecedented surge in demand for computing power 17 ; not only for edge nodes needed to low-latency solutions, but also for the broader computing capacity, i.e. the total volume of processing resources available, required to support fine-tuning of models and inference. Beyond AI, the adoption of cloud computing and other digital services continues to accelerate, further intensifying the pressure on available infrastructure. In this regard, the current investigations opened under the Digital Markets Act 18 in relation to the cloud computing sector are exploring the need and possibility to unlock opportunities and support fairness and contestability in the provision of cloud services. In addition, interested Member States are designing an Important Project of Common European Interest (IPCEI) focusing on the deployment of a Compute Infrastructure Continuum (CIC), namely a distributed and federated network of digital infrastructure, aiming to further increase compute capacity availability in Europe and enable functionalities, including but not limited to AI. In 2025, per Eurostat, EU business cloud uptake stood at 46.7% - still far from the 2030 target of 75%. As more European businesses adopt cloud and AI computing services, demand for data centres is therefore expected to rise further. In 2025, the EU’s computing capacity was estimated at approximately 12 GW 19 .

The EU continues to lag behind other regions in both the scale and ownership of digital infrastructure 20 . Despite comparable GDP levels, the EU accounted for only 20% of global data centre capacity in 2025, while the US held 42%. Although this capacity is expected to grow in the coming years, the gap relative to projected needs is also expected to widen. Market evidence points to tightening conditions for capacity expansion, with demand for colocation space in Europe exceeding new supply despite investments 21 . In 2025, demand for new data centre capacity in Europe reached a record of 854 MW, outstripping new supply for the third consecutive year. Across EU-27, the expansion of data centre capacity is therefore not able to keep up the pace with the rapidly growing demand 22 . Since 2022, average asking prices in European colocation markets have surged by 51% for 100 kW leases. 

While cloud and AI computing services can be technically delivered cross-border, regions with a low data centre presence are disadvantaged by the existing geographic imbalance in infrastructure deployment, as reflected in higher prices in regions with low data centre capacity 23 . Moreover, the lack of nearby computing capacity drives up latency, limiting the availability and quality of low-latency services, thus placing local end-users at a competitive disadvantage compared with regions that have better access to DC capacity 24 . Some of the identified key bottlenecks slowing down the deployment of computing capacity across the EU include regulatory fragmentation, permitting procedures, limited land availability, and increasingly, constraints of energy supply. The current regulatory environment remains fragmented, with different rules and permit requirements across Member States, creating uncertainty and delays for data centre operators. Permitting procedures are often lengthy and inconsistent, with multiple stakeholders and decentralised decision-making. Additionally, access to suitable land, affordable energy, and grid capacity is a significant challenge, particularly as energy prices in Europe are significantly higher than in other regions 25 .

Insufficient computing capacity in the EU could slow innovation and the diffusion of cloud and AI services, increase dependence on non-EU providers, and limit the ability of businesses and public services to meet growing demand for digital services. There is no direct quantification of the consequent direct impact on innovation, but literature suggests that AI adoption can generate significant gains in productivity meaning that any capacity gap would risk delaying or displacing these gains. This impact would not only concern AI deployment but also digital services that are heavily reliant on cloud infrastructure. Stakeholders, including Mistral AI, have warned that insufficient data capacity would become a roadblock for developing and applying AI in Europe. Over time, persistent disparities risk slowing digital transformation in affected member states, widening gaps in terms of digital adoption and deployment and thus undermining the competitiveness of the Digital Single Market.

Against this background, the Commission's Cloud and AI Development Act aims to triple EU data centre capacity within the next years, with a focus on sustainable infrastructure. Looking ahead, regular monitoring of this capacity would be beneficial, yet a significant challenge will be ensuring consistent measurement throughout the EU, as differences in definitions, metrics and scope currently complicate precise comparisons.

2.2.Supporting innovative companies with digital tools and resilient network

2.2.1.Connectivity

Connectivity is a fundamental enabler of the EU’s long-term competitiveness, determining how enterprises and public services can share data, innovate and deliver value. High-quality, secure cross-border connectivity enables companies to leverage cloud and AI services, scale up and collaborate in EU-wide value chains. Strong connectivity also improves the Union’s resilience and preparedness by ensuring continuity and flexible reconfiguration of business and public service operations, including critical ones, in the event of disruption.

According to the Digital Decade Eurobarometer 2026, 81% of Europeans consider it important for the EU to ensure access to high-speed internet for all EU citizens, while 83% of respondents think the EU should cooperate with Member States to build an independent European digital infrastructure (including broadband, 5G, cloud, semiconductors).

Modern connectivity is evolving from a combination of separate technologies - fixed, mobile, satellite networks or submarine cables - towards a more integrated ecosystem. To respond to the growing demand for critical, low-latency applications, such as connected and automated mobility (CAM) and human-machine interaction, increasing volumes of data 30  need to flow seamlessly across all these domains, so that disruptions in one domain do not undermine the performance and security of the system as a whole. Demands on cross border backbone connectivity are also increasing exponentially: currently, cables worldwide transmit over USD 10 trillion  in financial transactions on a daily basis 31 and capacity needs are expected to increase with AI and cloud applications. This evolution requires a global approach, spanning satellite, terrestrial and subsea infrastructures as part of a unified, resilient, and globally optimised system. Full fibre coverage, accelerated deployment of stand-alone 5G as a stepping stone to the development of 6G, and sufficient multi-orbit satellite capacity are the foundations for this approach.

Increasing attention is paid not only to coverage, but also to the security and resilience of networks, as well as to integrated connectivity systems, as means of ensuring redundant and reliable connectivity. The Council conclusions of 6 June 2025 32  call for a comprehensive approach to the development of a reliable and resilient network infrastructure, including via network diversification, interoperability and further deepening of the Single Market. They also called on the Commission to assess the possibility of a coordinated initiative for planning and developing a reliable and resilient network of digital infrastructures and capacities, including backbone terrestrial, submarine and satellite networks, across the Union and with international partner countries, for example by using the Trans-European Networks framework.

In their National Roadmaps, submitted in 2025, Member States reported devoting a significant portion (approximately 28%) of their measures’ total budget to gigabit fixed connectivity, amounting to EUR 80.9 billion (with EUR 56.6 billion coming from private sources). The 106 measures reported mainly focus on regulatory actions to facilitate network deployment, as well as financial support for non-viable and commercially unattractive areas and strategic infrastructure, including cross-border 5G corridors, submarine cables, and secure backbone networks. Around one third of the measures are dedicated to each of the two areas. In their roadmap adjustments, Member States maintained a strong emphasis on regulatory actions to facilitate network deployment.

In terms of 5G investments, Member States reported investing EUR 6 billion (with EUR 2.9 billion coming from non-public funds), which accounts for approximately 2% of the total budget of their measures. The 39 measures for 5G focus on spectrum management, as well as financial support for non-viable and commercially unattractive areas, and strategic parts of the network. There is equal emphasis on each of the two areas. In their roadmap adjustments, Member States placed a significant focus on increasing financial support for 5G networks.

However, recent territorial analyses indicate that improvements in connectivity do not translate uniformly into digital performance across regions. While infrastructure gaps persist in rural and peripheral areas, new divides are increasingly driven by differences in digital capabilities, usage patterns and local socio-economic conditions 33 .

Fixed access networks

The next five years will be characterised by a progressive shift from the current fixed access networks towards ubiquitous, full-fibre infrastructures (from fibre-to-the-premise to fibre-to-the-room), offering symmetrical multi-gigabit speeds, ultra-low latency, high reliability and much lower energy consumption. Fibre is increasingly viewed as a strategic, future-proof asset, capable of supporting data-intensive applications such as cloud and edge computing, AI, Augmented and virtual reality (AR/VR), remote healthcare, smart grids and cities, and industrial automation. The growing need for symmetrical capacity will require significant infrastructural upgrades towards large scale multi-gigabit architectures.

Despite these increasing stakes, the current pace of evolution in FTTH rollout remains insufficient.

The percentage of households with fibre connection rose by 4.9 percentage points, from 69.2% in 2024 to 74.1% in 2025, representing a year-on-year increase of 7.1%. According to the forecast along the baseline trajectory, 90.1% of the target is expected to be achieved by 2030 ( Figure 4 ). In 2025, the FTTP coverage stood at about 80% of the ideal value along the digital decade trajectory (74.1% instead of 94.0%). The full target - 100% of households covered - is forecast to be reached only in 2050 if no further actions are taken. Only 62.6% of households living in rural areas were reached by fibre in 2025, up from 58.8% in 2024 (+6.5%).

Figure 4: FTTP coverage in the EU. Historical data, Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

Take-up rates for high-speed fibre (FTTH) also vary significantly across Member States, remaining below 30% in some of them. Low adoption rates have reduced operators’ returns on investment, particularly for smaller fibre providers without an established customer base, consequently reducing incentives for further investments.

To accelerate the transition from copper to fibre, the Digital Networks Act proposal 34  sets an EU-wide copper switch-off objective for 2035. This would however be subject to strict conditions on fibre coverage (95%) and availability of comparable retail offers. The proposal is also accompanied by several safeguards and supporting regulatory measures, regarding, in particular, the deployment of the last part of the fibre network connecting the end-user.

The investment needed to achieve 100% FTTP coverage - the Digital Decade target for 2030 - is estimated at EUR 40 billion, or EUR 29 billion if 5G Fixed Wireless Access (FWA) is used for remote areas. Therefore, in order for the Digital Decade target to be achieved, it will be necessary to continue supporting fibre rollout, through the transition away from legacy copper networks.

Mobile networks

In the coming five to ten years, mobile networks are expected to evolve from basic 5G to advanced 5G standalone (5G SA) and to 6G, delivering much higher capacity, ultra-low latency, and native support for AI-driven and immersive applications. They will become increasingly software-defined, virtualised, and smart, as AI is embedded in network management, optimisation, and security. Mobile networks will also be tightly integrated with fixed fibre networks to support dense small-cell deployments and with non-terrestrial networks to ensure ubiquitous connectivity. Beyond consumer connectivity, future mobile networks will increasingly serve industrial, public-sector, and mission-critical use cases. This will enable automation, smart infrastructure, and real-time services, with a stronger focus on energy efficiency, resilience, and security as strategic requirements. Mobile data usage per mobile connection in Western Europe is projected to rise to 49 Gb per month by 2030, compared to approximately 15.3 Gb per month in 2024 35 . 

Convergence with fixed networks is expected to deepen, as fibre becomes essential for 5G SA/6G backhaul and fronthaul. Functional convergence with satellites is also progressing, with important implications for mobility and industry verticals and the connection of remote areas. 6G is expected to enable the seamless integration of terrestrial and non-terrestrial networks (NTNs), including satellite systems, into a unified architecture. This will support the scaling of direct-to-device (D2D) connectivity and enable ubiquitous, resilient and high-performance connectivity across both ground-based and space-based infrastructures. In this context, satellite connectivity is expected to become a native component of 6G networks, marking a shift from its traditional role in backhaul towards direct connectivity for end users.

Nevertheless, the EU’s progress towards this transition remains very slow, not when it comes to basic 5G coverage, but with regard to the 5G networks able to deliver transformative services.

Member States are indeed very close to full basic 5G coverage (96.8% of households) with substantial coverage increases, including in rural areas; all EU Member States have basic 5G household coverage above 90%, except Romania. However, basic 5G coverage is insufficient, given its technological limitations, to support the EU goals in the area of competitiveness.

Mid-band deployment - which is associated with high-capacity, high reliability and low latency - remains a bottleneck. 5G mid-band coverage in the 3.4-3.8 GHz spectrum band is substantially lower than general 5G coverage in many Member States (74.8% overall household coverage). Rural mid-band coverage is particularly weak. The EU27 average of 5G rural household coverage in the 3.4-3.8 GHz band remains at around 33%, indicating that progress is largely confined to urban areas.

This gap is closely linked to delays and modalities in the assignment of all three 5G pioneer bands. The process of authorising 5G pioneer spectrum across the EU has been lengthy, fragmented and insufficiently predictable, spanning more than a decade. In particular, most national auctions for the 3.4-3.8 GHz took place over a relatively long period, between 2017 and 2024. Member States that assigned spectrum earlier reached higher population coverage more quickly than later movers 36 . 

This gap in high-capacity 5G deployment limits the ability to deliver quality-assured services and advanced use cases, such as industrial automation, connected mobility and other mission-critical applications, and risks undermining Europe’s competitiveness in next-generation digital services.

Across Member States, 5G deployment remains predominantly based on Non-Standalone (NSA) architecture, which relies on existing 4G core 37 . 5G SA deployment (measured as 5G SA base stations as % of all mobile base stations) stands only at 20.9% in the EU, below US (36.2%), China (34.8%), Japan (26.3%) and South Korea (26.2%) 38 . Other sources estimate that only around 40% of the EU territory is covered by high capacity 5G SA, compared to 91% in North America and 45% in Asia-Pacific, highlighting once again a significant infrastructure gap.

Most EU countries launched NSA networks between 2019 and 2021, enabling early market rollout. However, the transition to 5G SA, which introduces a fully virtualised 5G core and enables advanced capabilities such as network slicing, ultra-low latency, and 5G SA private network services, i.e. capabilities that are important for competitiveness of EU industry, has been more gradual and uneven in the EU than in other advanced economies.

Europe is also significantly lagging behind in take-up as in “basic” 5G take-up (measured as a share of 5G SIM cards among all SIM cards), the EU stands at (28.1%), which again places the region behind the leading developed countries. The US (56.7%), Japan (54.7%), South Korea (41.0%) and China (39.6% all have higher rates 39 . As of Q4 2025, only around 2.8% of 5G users in Europe are connected via SA networks, compared with approximately 30% in the United States, and far behind India and China, which have reached around 52% and 81% respectively 40 .

The slow rollout of mid-band 5G, combined with the low pace of 5G SA adoption in Europe has direct repercussions for user experience. The EU27 average download speed of 69.9 Mbps remains below that of South Korea (162.2 Mbps), the US (129.3 Mbps), and China (100 Mbps) 41 . This performance gap is economically significant, as empirical evidence shows that improvements in mobile broadband capabilities - particularly speed - are associated with higher productivity and GDP 42 . Supporting evidence suggests that a 10% increase in mobile broadband speed is associated with a 0.2% increase in labour productivity in the subsequent period 43 . In this context, Europe’s comparatively weaker 5G performance may limit the realisation of potential productivity and economic growth gains.

In addition, the EU trails other regions in the deployment of private networks, which are early (industrial) adopters of 5G SA. Globally, 1489 private mobile network deployments have been identified, of which only 694 (47%) use 5G 44 . Manufacturing remains the leading sector for such networks, with 298 deployments recorded, of which around 60% include 5G. Asia-Pacific countries such as Japan and South Korea have progressed more rapidly, supported by more favourable policy frameworks and coordinated industrial strategies.

Relative to leading global markets, the EU remains less advanced in 5G investment intensity 45 . Its 51% 5G allocation share is solid (measured as % of all mobile investments), but below China (72%), South Korea (67%), the US (62%), and Japan (58%).

Achieving high-quality 5G SA coverage is estimated to require EUR 33.5 billion for network densification, with an additional EUR 26-79 billion needed to cover main transport paths 46 . The Digital Networks Act (DNA) proposal identifies disincentives and fragmentation in spectrum regulation as one of the main root causes of insufficient investments by European telecom operators in mobile networks. Other contributing factors include limited financial capacity and attractiveness for investors (itself due to low ARPUs, low predictability, etc.), low demand for advanced connectivity and unexploited economies of scale. For example, spectrum costs represent 7% of mobile service revenues and 35-40% of capital expenditure, reducing financial flexibility for 5G/6G investments. Additionally, fragmented and short-lived regulatory regimes act as disincentives.

A series of regulatory responses, essentially tackling the supply side, are offered in the DNA proposal, notably on spectrum policy. In particular, the DNA proposal simplifies and streamlines the regulatory framework that affects 5G and 6G rollout, with a view to reducing fragmentation across Member States and creating more predictable conditions. In doing so, it lays the groundwork for telecom innovation, particularly for emerging technologies such as 6G and satellite connectivity.

These proposed measures include unlimited spectrum licence duration by default and facilitated renewal procedures, affecting around 500 licences set to expire across Europe between 2025 and 2035, combined with safeguards such as periodic reviews, the possibility of revocation (e.g. in case of breach of conditions), and strong “use it or share it” obligations. They also include a pro-investment auction design, greater EU-level coordination through mandatory spectrum scrutiny and harmonised authorisation conditions (i.e. ex-ante Spectrum Single Market mechanism to ensure that auctions align with the objectives of the DNA) as well as faster and more predictable authorisation of future 6G spectrum to enable timely deployment. These measures are accompanied by other proposed measures on authorisation and governance, designed to facilitate larger scale operations and unlock the full potential of the single market.

The DNA also introduces a Union radio spectrum strategy to guide long-term spectrum planning, identify future needs, and ensure the availability of spectrum for key services and technologies.

It will, however, remain important to sustain public intervention in mobile access networks, including on supply side measures (i.e. public support to 5G/6G rollout) targeting areas of market failure in order to meet the Digital Decade targets, support cohesion and ensure that the Union’s full industrial potential is fully exploited. Further actions remain necessary to support the ongoing implementation of the EU 5G Cybersecurity Toolbox and once adopted, to facilitate alignment with the trusted ICT supply chain security framework pursuant to the revised Cybersecurity Act.

Moreover, in order to address the root causes identified above, regulatory and funding measures may be complemented by demand-side stimulation, with a particular focus on enabling innovative business models, including the bundling of infrastructure deployment with edge cloud and AI integration and with concrete use cases.

Since 2021, the digital part of the Connecting Europe Facility programme (CEF Digital) has co-funded 5G deployments integrated with edge-cloud computing and enabling innovative use cases such as remote surgery, virtual reality for learning, drone-based monitoring and more. So far, a total of EUR 327 million were invested in 78 projects, including 47 projects for 5G Smart Communities and 31 projects for transport corridors. This funding has helped pave the way for the future development of vertical use cases for sectors considered as strategic for the economy.

In 2026, the Commission unveiled EURO-3C, a EUR 75 million project meant to develop the EU’s first large-scale federated Telco-Edge-Cloud infrastructure 47 . Financed via Horizon Europe, this landmark project, which brings together over 70 partners, will showcase the EU's ability to deliver cutting-edge digital services entirely through its own connectivity infrastructure, reducing reliance on third country providers. Telco-edge-cloud combines telecommunication networks, edge computing and cloud infrastructure into a single, integrated platform, bringing high speed, secure computing power closer to end-users. 

EU’s leadership in 6G will not be determined solely by leadership in radio technologies, but by its capacity across the converged digital communications stack, from advanced semiconductors to AI-driven network orchestration and cloud-edge integration.

The EU approaches the transition to 6G with a mix of structural strengths and growing dependencies 48 . It benefits from globally competitive vendors, strong radio access network (RAN) engineering capabilities, and recognised leadership in energy-efficient networks, supported by a coordinated research and standardisation framework, notably through the Smart Networks and Services Joint Undertaking (SNS JU) as well as large national 6G initiatives to promote European capacities in 6G and related technologies in a number of Member States (in particular Germany, Ireland, Spain, France, Italy, Netherlands, Finland, Sweden) 49 . These assets strengthen the EU’s ability to influence global standards and ensure interoperability. However, challenges persist, including weaknesses in hyperscalers cloud services and AI development, reliance on external semiconductor supply chains, risks of diminished value capture in increasingly software-driven architectures, and exposure to geopolitical tensions and market fragmentation.

As value shifts toward cloud management software and AI orchestration, the EU risks losing ground in higher-margin segments if these capabilities remain externally dominated.

In conclusion, public investment - covering inter alia R&I, supply chain, network deployment, and fostering the take-up of advanced services - must continue to complement the new rules put in place with the proposed Digital Network Act, once adopted. Based on the lessons learnt from the current MFF, it will be essential to support 5G and 6G across the innovation journey, coupling network deployments with use cases and with the necessary edge, cloud and computing resources.

Satellite connectivity

Satellite systems provide broad regional and global coverage and, by their nature, can support a pan-European (or global) reach, unlike terrestrial mobile networks, which remain bounded by national deployment. As direct-to-device (D2D) connectivity is emerging, complementing terrestrial mobile services, as well as machine-to-machine services, including in underserved areas, the current EU framework - based on national authorisation regimes and national spectrum allocation - creates barriers to the provision of pan-European services. It also forces satellite operators to comply with divergent national rules and conditions across Member States in which they operate.

In particular, Low Earth Orbit (LEO) constellations supporting hybrid terrestrial-satellite systems or Non-Terrestrial Networks (NTN) are emerging as the modern equivalent of traditional mobile communication towers. They enable D2D connectivity and are expected to become an essential component of future 6G networks. Integration with terrestrial mobile networks (5G SA/6G) is expected to become central to future communications systems, enhancing resilience and ubiquity of connectivity. D2D satellite connectivity is rapidly emerging and may play a strategic role for mobile network operators (MNOs) and smartphone equipment manufacturers (OEMs). It presents a valuable opportunity to stand out in a saturated market, improve customer retention, and build long-term value. A global survey by Analysis 50 , covering 18 500 respondents across 18 countries, highlights strong demand for D2D satellite messaging, showing that an early adoption of satellite D2D may be leveraged to attract and retain subscribers. Notably, 82% of subscribers considering switching providers within the next six months expressed interest in such services while 30% said they would be willing to pay for them.

Satellite infrastructure is also critical for ensuring equitable access to high-speed internet, especially in underserved and remote areas, while also strengthening the EU’s capabilities in critical communications, including emergency and defence services. Although Europe was once a frontrunner in satellite communications, it has been slower to anticipate the innovation potential of this market and to invest in LEO constellation deployment and now lags behind the US and China. This increases the risk that the EU becomes dependent on non-EU providers of this critical infrastructure, with implications for both competitiveness and digital sovereignty, especially in a context of geostrategic uncertainty. According to a European Commission study 51 on Mobile Satellite Services in the 2 GHz band, as of March 2025 the EU has only 773 LEO satellites launched and 3 120 planned, compared with 220 launched and 27 198 planned in China and 7 633 launched and 33 397 planned in the US.

Looking ahead, the number of authorisations is expected to increase significantly, raising compliance costs for operators and enforcement costs for authorities. The coexistence of twenty-seven national authorisation regimes also contributes to coverage gaps and slows the rollout of pan-European satellite services.

The proposed DNA and MSS (Mobile Satellite Services) 52 Regulations introduce a single EU-level authorisation for satellite services ensuring EU-wide access to spectrum under harmonised conditions. This framework aims to create a level playing field, enable European operators to scale up, and support the development of innovative satellite services, such as D2D connectivity, which are increasingly critical for EU security and resilience. It also aims to strengthen the global competitiveness of EU operators and is expected to unlock further investment in satellite infrastructure. However, given the capital-intensive nature of satellite deployments and prolonged revenue realisation timelines, structured public-private partnerships, and strategic Union initiatives, remain key to accelerating this transformation and securing independent European infrastructure capabilities.

Backbone networks

The growing traffic in access networks, combined with rising data flow from and to edge, cloud, AI capacities and Content Delivery Networks, will require massive scaling of backbone networks. This will rely on advanced optical technologies such as coherent transmission, higher-order modulation, and open line systems, enabling multi-terabit capacities per fibre. Backbone networks will increasingly interconnect distributed data centres and edge nodes, reduce latency and support real-time and mission-critical services.

Backbone architectures will also become more software-defined and automated, as resilience, security, and redundancy become increasingly strategic priorities. This includes route diversification, protection against physical and cyber threats, and greater focus on submarine and cross-border terrestrial links as well as related deployment, maintenance and repair capacities (e.g. multi-purpose, modular vessels). The migration process to Post Quantum Cryptography shall safeguard the continuity of services, availability of data, and use of applications, and require coordinated action also across different types of networks, given the many interdependencies and interfaces. Quantum communications technologies will be integrated into critical backbone networks to support highly secured, mission-critical applications (e.g. QKD between Member States or banking companies).

The rollout of future backbone networks will not only be driven by cross-border and intercontinental connectivity needs but also by the deployment of data centres and computing capacities in geostrategic and areas where renewable energy is abundant, located close to clean power sources (e.g. solar or hydropower plants).

In particular, the total value of submarine communication cables’ global sales was estimated at USD 3.8 bn in 2024 and is expected to reach at least USD 7 bn by 2034 53 . In November 2024, Analysys Mason predicted even higher figures, up to USD 10 bn in 2029.

While the EU has good presence and strengths in the submarine/backbone networks market, it faces strong and increasing pressure from global competitors. Current market trends show a steady decline in European investment, leading to the exit of European actors and a loss of EU market share 54 . Meanwhile, the US continues to invest in high-capacity backbone networks, driven largely by hyperscalers, content providers, and cloud companies.

Between 2019 and 2023 the amount of international submarine cable capacity deployed by GAFAM (Google, Apple, Facebook, Amazon, Microsoft) quadrupled. Today, Google, Meta, and Amazon own 59 international submarine cables , up from just 20 in 2017, meaning that the bulk of capacity is now held by non-EU controlled companies. 

In terms of cable manufacturing and deployment, the main US player, SubCom, has the greatest market share worldwide, followed by French-owned Alcatel Submarine Networks (ASN) and Japan’s NEC. China’s Huawei Marine Networks (HMN) is lagging behind but is gaining market share rapidly. Additionally, the US and Japan have recently announced massive investments in deployment capacities (including icebreaker vessels). The EU Risk Assessment published by the Commission in October 2025 provides a full analysis of the market and stakeholder ecosystem 55 . 

Given their criticality and of the rapidly evolving markets that determine network topography, capacity, but also resilience, security and control, the EU has paid increasing attention to backbone networks and in particular submarine cables, proposing a comprehensive policy approach, including regulatory and funding measures.

On the policy side, the EU Action Plan on Cable Security has outlined a series of coordinated actions to address risks and enhance the security and resilience of data and power submarine cable infrastructures, across the full resilience cycle: prevention, detection, response and recovery, and deterrence 56 . To support the implementation of the 2024 Cable Recommendation 57 and 2025 Action Plan, the Commission set up the Submarine Cable Infrastructures Expert Group and published in October 2025, its EU risk assessment (based on threats, vulnerabilities and dependencies), including mapping and stress test guidance on the security and resilience of EU submarine cable infrastructures 58 .

Furthermore, on 5 February 2026 the Commission published the Cable Security Toolbox, which recommends a set of mitigation measures to address the identified risk scenarios, as well as a list of Cable Projects of European Interest (CPEIs), i.e., areas to be prioritised for public funding 59 . 

On the funding side, to date the EU invested over EUR 600 million in more than 70 Digital Global Gateways projects through the Connecting Europe Facility Digital programme (CEF Digital). While a number of satellite-terrestrial links and cross border terrestrial backbone projects have been funded, the vast majority of the funding was dedicated to submarine cable projects (59 projects for EUR 548 million). These projects are enabling significant improvements in the coverage of islands and remote territories, reduce vulnerabilities and risks, while increasing the overall resilience and redundancy of connectivity systems within the EU and linking the EU to the world, contributing to strategic objectives such as the connection of Europe to Asia through the Arctic, including Greenland’s and other areas identified in the CPEI list.

In line with the EU Action Plan on Cable Security, the Commission amended the CEF Digital Work Programme , in February 2026, allocating EUR 347 million to fund strategic submarine cable projects 60 . These calls will support the CPEIs, including to enhance the EU's cable repair capacity, and equip submarine cables with smart capabilities.

In 2026, two funding calls worth EUR 60 million will support cable repair modules, alongside a separate EUR 20 million call for SMART cable system equipment. These are sensors and monitoring components integrated into submarine telecommunications infrastructure to gather real-time ocean and seismic data. Additionally, two calls for new CPEI cables are planned for 2026 and 2027, with a total budget of EUR 267 million.

However, while CEF has already intervened in the CPEI areas, and will continue to do so in an increasingly focused manner, the remaining funding capacity of CEF Digital remains modest compared to the total investments needs. This gap is also reflected in the increasingly high oversubscription rates in CEF Global Gateway calls 61 . Cost estimates for deploying CPEI projects exceed EUR 10 billion 62 . Other studies suggest that at least a threefold increase of the current level of funding per annum for submarine cables (i.e. EUR 200 million) will be needed to maintain the current level of European ownership in the cable market and give greater scope and ambition to address market fragilities and critical capability gaps.

It is therefore important that support continues for the deployment, upgrade and maintenance of submarine cables to preserve competition and foster increased reliability and security standards. Deploying submarine cables and equipping them with advanced security monitoring and rapid-repair capabilities remains vital and requires continued public funding in the next MFF. In line with the Cable Security Toolbox, this intervention must focus on routes that are strategic for the EU (CPEI areas) and reduce reliance on non-EU suppliers.

In parallel, the DNA introduces new provisions concerning the resilience of electronic communications networks and services. These include cooperation and coordination of resilience and preparedness actions, data collection, early warning, networks’ resilience mapping and networks’ capabilities to ensure the redundancy through different types of networks’ backups. The purpose is to identify potential bottlenecks where resilience-enhancing measures are needed at Union level, including strategic investments to support redundancy, in particular, for trans-European digital networks.

In this context, the newly proposed Office for Digital Networks (ODN) will be tasked, once the DNA proposal is adopted, with ensuring a coherent, cross-border approach for electronic communications networks and services, among others by preparing the Union Preparedness Plan for Digital Infrastructures, which will include a comprehensive assessment and an overview of network topology at Union level, identify route diversification, potential bottlenecks or points of failure and areas where resilience-related measures, such as strategic investments to support redundancy, are needed.

Such information, especially on integrated connectivity covering a wide range of terrestrial and non-terrestrial communications networks, can support a wider analysis, complementing the one which led to the Cable Projects of European Interest, and thereby ensuring an updated, integrated prioritisation of critical investments in strategic backbone networks. Building on broader orientation set out in the Council conclusions on Connectivity of 6 June 2025, the Council also invited the Commission to assess the possibility of a coordinated initiative for planning and developing a reliable and resilient network of digital infrastructures and capacities, including backbone terrestrial, submarine and satellite networks, across the Union and with international partner countries.

2.2.2.Digitalisation of SMEs and uptake of digital technologies by EU enterprises: AI, Cloud and data analytics

While building cutting-edge capacity and infrastructure is a necessary condition for digital competitiveness, the diffusion and deployment of advanced technologies across businesses is key to ensure that technological leadership translates into competitiveness gains. 

The broad diffusion of technologies across firms, sectors and regions is therefore a second fundamental challenge facing the EU, and a key driver for productivity growth and economic transformation. The data on cloud, data analytics and AI up-take confirm that headline adoption figures are moving in the right direction, but progress is too slow relative to 2030 targets, and SMEs consistently lag behind large enterprises.

The percentage of SMEs with at least a basic level of digital intensity, according to version III of the Digital Intensity Index, rose by 13.5 percentage points in two years, from 57.90% in 2023 to 71.38% in 2025. This reflects a year-on-year increase of 11.0%. 

According to the forecast along the baseline trajectory, the target is expected to be achieved and surpassed by 2030, with an estimated 92.08% of SMEs having at least a basic value of the digital intensity index by then ( Figure 5 ). The revised baseline trajectory is fully in line with the ideal trajectory. In 2025, the observed and ideal values overlap almost perfectly. 

In their National Roadmaps, 165 measures corresponding to a total of EUR 48.4 billion were reported as supporting the digitalisation of SMEs. The measures primarily focus on facilitating the uptake and deployment of digital technologies, as well as strengthening the broader ecosystem through activities such as information sharing, knowledge exchange and collaboration on digital technologies.

Figure 5: Digital Intensity Index ver. III. Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

The percentage of enterprises using at least one of the three technologies (Cloud, Data Analytics or AI) rose by 8.5 percentage points, from 54.7% in 2023 to 63.20% in 2025. This represents a substantial year-on-year increase of 7.5%. 

According to the forecast along the revised baseline trajectory, 95% of the target is expected to be achieved by 2030, with 71.5% of enterprises expected to use at least one of the three technologies by then ( Figure 6 ). In 2025, the value of this KPI aligns perfectly with the ideal trajectory, reaching a value of 63.2%. The target of 75% of enterprises using at least one of the three technologies is expected to be reached in 2033 if no further actions are taken. 

Figure 6: Percentage of enterprises using AI or cloud computing or data analytics in the EU. Historical data, Digital Decade (DD) trajectory and baseline trajectory towards 2030 (same speed of diffusion parameter as in cloud baseline trajectory).

Artificial Intelligence 

According to Eurostat data, the percentage of enterprises using AI technologies rose by 6.5 percentage points in one year, from 13.5% in 2024 to almost 20.0% in 2025. This represents a substantial year-on-year increase of 48%, however lower than the 67% year-on-year increase of 2024.

In EU companies, the most widely used AI technologies are those for analysing written language (11.8%), which also saw the strongest increase since 2024. They are followed by AI for generating images, videos or audio (9.5%), producing written or spoken language (8.8%), and converting speech into machine-readable format (7.2%). According to the forecast along the baseline trajectory, 72% of the target is expected to be achieved by 2030, with 54.28% of enterprises expected to adopt AI by then ( Figure 7 ). In 2025, this KPI reached a level slightly above half of the ideal value defined by the ideal trajectory (at almost 20.0% instead of 39.1%). The target of 75% of enterprises using AI is expected to be reached not earlier than 2035 if no further actions are taken. 

In their National Roadmaps, Member States reported investments of EUR 10.9 billion to support the uptake of AI, cloud or data analytics - representing approximately 3.8% of the total budget across roadmaps and covering 199 measures in total. Among these, roughly 34 measures specifically target AI, accounting for EUR 1.4 billion. Measures supporting the uptake of AI, cloud and data analytics are evenly distributed across measures to enhance the ecosystems and knowledge exchange, establish enabling framework conditions and develop capabilities across these technologies. However, AI-specific measures place a stronger emphasis on building AI capabilities. This focus is also reflected in Member States’ roadmap adjustments.

Figure 7: Percentage of enterprises using AI in the EU. Historical data, Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

These figures indicate that, despite strong recent progress, including on policy making at EU and national level, the EU remains below the Digital Decade target. Reaching 75% of enterprises by 2030 will require targeted support, especially for SMEs and for sectors and regions that lag behind. They also show that large firms still adopt AI much faster than SMEs, even if smaller firms are accelerating.

For European SMEs, the evidence suggests that the main obstacle is no longer basic awareness of AI but the difficulty of turning interest and pilots into broad business deployment. Eurostat’s latest statistics support this analysis. Among EU enterprises that had considered using AI but did not adopt it, the main reasons reported were lack of relevant expertise (70,89%), lack of clarity about legal consequences (52,52%) and concerns regarding violation of data protection and privacy (48,83%). For many firms the issue is no longer whether AI matters, but whether they can implement it safely, lawfully and effectively.

Concretely, for SMEs, the most recurrent bottlenecks are fragmented and poor-quality data, weak interoperability systems, lack of in-house skills, uncertain returns on investments and difficulties in integrating AI in core business processes. These barriers are compounded by organisational and compliance difficulties, with many SMEs not having the internal capacity to identify viable use cases, adapt workflows, assess risks, or manage legal uncertainty around data use and AI deployment.

The main challenge is therefore one of scale and broad-based diffusion across industrial sectors with particular focus on the strategically important ones.

For instance, the OECD Study 63  - indicates that in manufacturing, AI uptake is still “modest and highly fragmented”: the share of manufacturing enterprises using at least one AI technology rose from 7% in 2021 to 11% in 2024, but the technology is still not well integrated into core production processes. The most common uses are still language-based and administrative tasks, while uptake remains lower for functions more directly tied to production, such as image recognition, robotic process automation and machine-learning-based optimisation.

In public transport and logistics, the report finds that many deployments remain at pilot or early implementation stage and that SMEs often struggle to keep pace because of limited access to expertise, infrastructure and funding.

Across sectors, the OECD repeatedly identifies fragmented datasets, poor interoperability, integration into legacy systems, compliance complexity and weak internal skills as persistent constraints on scaling.

For SMEs, de-risking adoption has become essential: they need trusted intermediaries, simpler access to expertise, and support that combines technology, compliance and business transformation rather than addressing these issues separately. This is consistent with the direction already taken by the Apply AI Strategy, which includes the role of the European Digital Innovation Hubs (EDIHs) as Experience Centres for AI and as key partners helping SMEs to “test before invest”, identify funding for their AI projects or network within the AI ecosystem of AI factories, Testing and Experimentation Facilities (TEFs) and future national sandboxes.

At the national level, by the turn of 2026, the vast majority of Member States have already adopted and/or revised their own national AI strategies. However, the national implementation remains uneven in terms of dedicated budget allocated, measurable KPIs or regular evaluation.

For instance, funding AI at national level is highly uneven and often difficult to isolate because many AI measures are combined with wider digitalisation plans. This creates a divide between countries with operational governance, measurable follow-up and financing mechanisms for AI and those where AI remains embedded in broader agendas without a strong delivery architecture. Moreover, some member states do not use well-defined KPIs for AI strategies, using instead broad digital transformation frameworks.

At European level, stronger coordination is needed to connect strategy, infrastructure and deployment. The AI Continent Action Plan provides the upper-level framework for the infrastructure and enabling layer, notably through AI factories, high-quality data access and compute capacity. The Apply AI Strategy is particularly important for the uptake objective, as it is explicitly the EU’s overarching sectoral AI strategy and is designed to boost adoption and innovation across Europe, especially among SMEs. In this architecture, EDIHs are meant to function increasingly as AI Experience Centres linked to AI Factories, Testing and Experimentation Facilities (TEFs) and regulatory sandboxes, while the AI Skills Academy supports the workforce dimension.

For AI uptake in EU companies, this EU-level coordination is critical: Member States on their own cannot efficiently provide the full combination of compute, high-quality data, regulatory support, testing infrastructure, skills and cross-border market scale needed for rapid AI uptake. Related to the regulatory framework, national market surveillance authorities, which each Member State is required to designate under the AI Act, are responsible for enforcing compliance with the rules applicable to AI systems, including by proposing joint investigations with the European Commission. To the best of our knowledge, 10 Member States (Denmark, Finland, Hungary, Ireland, Italy, Latvia, Lithuania, Malta, Slovenia and Slovakia) have designated their national market surveillance authorities 64 . The timely establishment of these authorities, together with sufficient technical, financial and human resources, will be critical to ensuring robust oversight and effective enforcement of the AI regulatory framework. 

The urgency is clear when current performance is compared with EU ambitions. Reaching the AI targets will depend on aligning reforms and investments across all levels of responsibility (European, national, regional and local) around a key objective: making AI adoption easier, safer and more affordable for EU companies, especially SMEs.

Cloud computing services

The percentage of enterprises using sophisticated or intermediate cloud computing services rose by 7.7 percentage points, from 38.97% in 2023 to 46.69% in 2025. This represents a year-on-year increase rate of 9.5%. According to the forecast along the updated baseline trajectory, approximately 81% of the target is expected to be achieved by 2030, with 60.6% of enterprises estimated to adopt sophisticated or intermediate cloud services by then ( Figure 8 ). The target is expected to be reached not earlier than 2040, if no further actions are taken. 

Figure 8: Percentage of enterprises using intermediate and sophisticated cloud computing services in the EU. Historical data, Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

Cloud computing remains critical for business digitalisation and is becoming an increasingly relevant enabler for AI deployment and uptake. Around 53% of enterprises used paid cloud computing services in 2025, an increase of around 8 percentage points with respect to 2023 65 . However, the most advanced end of this adoption is less widespread, with approximately 41% of enterprises buying at least one sophisticated cloud service. This demonstrates that adoption is growing but most firms are still using basic services. Cloud uptake is advancing but too slowly relative to the ambition of the 2030 target, which makes this recent acceleration not sufficient on its own.

One of the factors holding back progress is the persistent SME gap. In 2025, 46% of SMEs used cloud services, compared with 60% for enterprise 50-249 employees and 78% of large enterprises. The comparison across countries is equally significant, with Finland’s SMEs reaching 72% of average cloud uptake while Romania, Greece and Bulgaria remaining below 25%.

To close this gap, policy support needs to target late adopters, especially SMEs, through practical support and advisory services. Stronger enforcement and implementation of the switching rules under the Data Act, entered into force in September 2025, could also help firms switch providers more easily and adopt cloud with lower lock-in risks. Progress towards the target will depend both on demand-side incentives, especially for these lagging groups and on whether the EU can build a sufficient, secure and sustainable computing infrastructure to accompany this transition. Without this combination, cloud uptake may continue to rise, but as shown above, too slowly to achieve the 75% target by 2030.

Finally, cloud can be considered an essential enabler for the uptake of data analytics and AI services, detailed below. For this potential to materialise, more firms will have to move from basic cloud consumption to more advanced data and AI-related uses.

Data Analytics

The share of enterprises using analytics rose from 33.25% in 2023 to 39.85% in 2025, an increase of 6.6 percentage points. Despite this progress, the EU remains off track to meet the Digital Decade target by 2030: under the revised baseline trajectory, only 51.9% of enterprises are expected to use data analytics by 2030, corresponding to around 69% of the target, and the target would be reached only by 2047 in the absence of additional action. 

Figure 9: Share of enterprises using Data Analytics in the EU. Historical data, Digital Decade (DD) trajectory and baseline trajectory towards 2030.

The EU has laid important foundations for the data economy through the European Strategy for Data and related key legislative instruments such as the Data Governance Act, the Data Act, and the Open Data Directive. However, significant barriers continue to limit the wider uptake of data analytics by enterprises, especially SMEs.

A first major challenge is data scarcity and limited access to high-quality datasets for data analytics and AI development. Valuable datasets remain siloed within organisations, fragmented across sectors, or difficult to access across Member States’ borders. This particularly affects start-ups and SMEs, which often lack access to sufficiently large, high-quality datasets and the computing resources needed to use them effectively. As global competition intensifies, broader and more reliable access to quality data becomes increasingly important for innovation and competitiveness.

A second challenge is regulatory complexity and fragmentation. The EU has developed a broad framework governing data access, sharing and protection, but interaction between horizontal and sectoral rules, combined with uneven implementation across Member States, can generate legal uncertainty and increase compliance costs. This is particularly burdensome for SMEs and emerging data intermediaries and can slow the scaling-up of data-driven business models and European data ecosystems.

A third key challenge is the strategic and geopolitical dimension of data. Data has become a key economic and security asset. Jurisdictional issues remain a critical concern. Data held by companies subject to non-EU jurisdictions may be accessed by foreign authorities, including under extraterritorial legal frameworks such as the US Cloud Act. This raises questions about effective control, legal certainty and the enforceability of EU rules, particularly for sensitive or strategic datasets. European firms must be able to benefit from secure cross-border data flows and access to global data resources, while the EU must also protect sensitive and strategic datasets and ensure that data governance remains consistent with European values.

Addressing these barriers will require coordinated reforms and investment. Priority actions include scaling up access to high-quality data for AI and innovation at large, notably through the further deployment of common European data spaces and strengthening interoperability through open and modular solutions. Linking data spaces with data labs and AI factories could help transform Europe’s data assets into resources for trustworthy AI and advanced analytics.

Further efforts are needed to simplify and modernise the EU data regulatory framework, making it clearer, more innovation-friendly, and less burdensome for businesses, especially SMEs. In this context, the proposed Digital Omnibus aims to streamline existing data legislation by reducing overlaps, updating certain rules (including on privacy and data use), and clarifying the implementation of key instruments such as the Data Act, while preserving their core objectives.

Strengthening EU data sovereignty and fair international data flows also remains essential. This requires enabling cross-border access to data, improving access to larger and more diverse datasets for businesses, and addressing remaining barriers to data sharing. Such measures are critical to support the development of advanced data analytics and AI applications and to enhance the global competitiveness of European companies.

2.2.3.Open Source

Open source is a strategic enabler of the digital transformation of the European Union and the Digital Decade goals as it underpins most modern digital systems and directly affects Europe’s competitiveness, resilience and technological sovereignty. Open source makes up 70 to 90% of all code in the digital economy 66 . A 2024 Harvard Business School study 67  estimated the demand-side value of open-source software at USD 8.8 trillion and found that firms would need to spend 3.5 times more on software if open source did not exist. The EU starts from a position of strength, with more than 3 million open-source contributors, more than 500 for-profit open-source companies and substantial EU support to open-source actions across cloud, AI, cybersecurity, internet technologies and chips, estimated at around EUR 800 million in the current MFF. Yet these strengths are still not converted into sufficient market scale, stewardship capacity or control over critical parts of the digital stack.

The main hurdles are structural. First, the EU still struggles to move from research and community development to adoption at scale. Many promising projects lack financing for integration into real-world environments, user experience improvements, security hardening, performance testing, legal compliance and commercial deployment. Second, maintenance remains a major market challenge. Critical open-source components are often widely used but maintained with fragile resources, creating both security and continuity risks. Third, access to scale-up capital remains weak, particularly for European SMEs and mid-caps building open-source business models. Fourth, public procurement frameworks still tend to favour incumbent proprietary suppliers by focusing on short-term pricing, product bundles and vendor-specific features rather than lifecycle cost, interoperability, exit costs and strategic control. Fifth, the EU still lacks sufficiently strong stewardship, governance and trust organisations that can carry out assessments, security attestations to make open software easier to buy and deploy in regulated or mission-critical environments. These challenges are amplified by fragmentation across Member States and by continued dependence on non-EU digital infrastructures and services, including software repositories, code hosting platforms, cloud execution environments and trust services that remain outside EU jurisdiction.

Investments and reforms are therefore needed to address this situation. Public procurement should better recognise open standards, interoperability, reusability and total cost of ownership, including switching and lock-in costs so that open software can compete on equal terms in tenders. Public administrations should be supported with common guidance, model clauses and skills to be able to migrate to open-source solutions with confidence. The EU should also strengthen trust and adoption mechanisms by developing shared assessments and security assurances, promoting common catalogues, rolling out interoperability frameworks and reusing open digital building blocks across borders. The open-source licensing requirement for the EU Digital Identity Wallet and the launch of the Digital Commons EDIC show that the EU already has practical anchors on which to build a more coherent adoption strategy.

Investment needs should be organised across the full lifecycle of open software. This includes R&I funding for strategic open technologies, but also dedicated uptake and support, multiannual maintenance funding for critical components, and financing for European stewardship structures. In this respect, the proposed European Competitiveness Fund is highly relevant. With a budget of EUR 234 billion overall and instruments designed to combine guarantees, financial instruments, blending, support to start-ups and scale-ups, and advisory services, it could help address the current gap between technical excellence and market deployment. For open software, priorities should include supporting deployment-ready sovereign solutions in strategic domains; de-risking adoption by public administrations and SMEs; financing maintenance and security of critical dependencies; and mobilising equity, quasi-equity, guarantees and advisory support for European open-source firms and integrators. This would make open software easier, safer and more attractive to deploy at scale across the Single Market.

EU Open-Source Strategy will review the current state of play and put forward a set of actions to be rolled out in the coming years by the public sector at European and Member State level, as well as the private sector and the open-source communities themselves.

2.2.4.Unicorns

The number of unicorns rose by 30 units (net increase), from 294 in 2024 to 324 in 2025, corresponding to 64.8% of the EU target. This represents a significant year-on-year increase of 10.2%. According to the forecast along the revised linear baseline trajectory, 88.0% of the target is expected to be achieved by 2030, with 440 unicorns expected to be active in the EU by then ( Figure 10 ). The target of 500 unicorns, set in the 2023 Communication on EU-level trajectories, is expected to be reached in 2033 if no further actions are taken.

Figure 10: Number of unicorns in the EU. Historical data and revised baseline trajectory.

Newly released data confirms the post-2020 linear trend but also point to an acceleration in the creation of new unicorns. In 2025 was almost twice as much than those created in 2024, the EU recorded 30 new net unicorns in 2025 compared to 16 the previous year. The yearly rate of progression has increased from 5.8% in 2024 to 10.2% in 2025 ( Figure 11 ).

Figure 11: Number of unicorns in the EU by year: annual net increase (births-deaths) and relative annual progress since 2009.

Comparing the performance of the EU’s startup ecosystem with that of its main global peers provides useful context for assessing its relative strengths and areas where further progress may be possible. Analysing the time series of unicorn creation since 2008 helps to place recent performance in a longer-term perspective and to compare the evolution of the EU’s startup ecosystem with that of its main competitors - Canada, China, Israel, Japan, South Korea, United Kingdom and United States ( Figure 12 ). 

Figure 12: Number of Unicorns in the EU and its main competitors from 2008 to 2025.

The European Union expanded its unicorn base from 3 in 2008 to 324 in 2025, with growth accelerating after 2015 and again after 2020. Despite this progress, the EU continues to lag well behind the United States, which reached 1 886 unicorns by 2025 - almost six times the EU total - highlighting persistent gaps in market scale and late-stage financing. China number of unicorns was 40% higher than that of the EU in 2025 (452), while the United Kingdom alone accounts for more than half of the EU’s unicorn total. Meanwhile, competitors such as Israel and Canada have scaled rapidly within large or well-integrated domestic markets.

Normalising the unicorn ecosystem by economic size confirms the EU’s underperformance ( Figure 13 ). Measured as the number of unicorns per USD 100 billion of GDP (2025 values)21, the European Union records 1.54, well below Israel the top performer (7.04), the United States (6.16), the United Kingdom (4.45), Canada (3.24), China (2.33), and South Korea (1.99), while remaining above Japan (0.37). The gap with leading ecosystems such as the U.S. and Israel is sizeable, while the distance to countries such as the UK and Canada, though smaller, also remains noticeable. With an economy of around two-thirds the size of the U.S., the EU generates fewer than one-fourth as many unicorns per unit of GDP. It is also worth noting that the strong performance of the U.S. is highly concentrated geographically. California plays a disproportionate role in driving the U.S. unicorn ecosystem. Despite accounting for around 13% of U.S. GDP (40.5k billion in 2024, most recent available year22), California recorded 892 unicorns in 2025, representing almost half of all U.S. unicorns (1 886). This concentration highlights the importance of dense venture capital markets, leading technology hubs, and strong innovation networks, in shaping ecosystem performance.

Overall, the international benchmarking analysis points to a significant margin for improvement in the EU’s ability to convert economic capacity into large-scale, high-growth firms if it is to narrow the performance gap with its global peers.

Tracking the outflow of EU-founded unicorns’ relocation to third countries provides an important indication of the EU’s capacity to retain high-growth innovative firms and to scale them within its own market22. The data show a gradual improvement in the EU’s capacity to retain innovative scale-ups with respect to the U.K. and the U.S. over the past decade ( Figure 14 ). Between 2016 and 2019, around one in five EU-founded unicorns relocated to the US or the UK, with percentages fluctuating between 21% and 22%. This indicates a relatively stable but significant outflow during that period. From 2020 onwards, the percentage of relocating companies declined more clearly and reached 16% in 2025. Other potential destination countries were also examined (Canada, China, Israel, Japan, South Korea), but no cases of relocation of EU-founded unicorns were identified outside the U.K. and the U.S. over the period considered.

The trend suggests a moderate but consistent strengthening of the EU’s ability to retain its innovative scale-ups. While relocation remains a structural challenge, with roughly one in six unicorns still moving abroad, the outflow has decreased by around five to six percentage points compared to the 2016-2018 peak.

Figure 14: Percentage of unicorns founded in the EU in the past decade that later relocated to either the United Kingdom or the United States.

Challenges facing EU startups on the path to unicorn status

In the EU, the path from promising startup to unicorn presents obstacles that are often more pronounced than in other competing regions. Three non-exhaustive challenges stand out in particular: the scale-up funding gap, weak spinoff creation and tech transfer policies, as well as the difficulty startups face in accessing public procurement markets.

The most glaring obstacle is the lack of late-stage capital for startups seeking to raise EUR 100 million or more. While early-stage funding has improved, EU startups frequently hit a wall when they need large-scale investment to expand globally. Fund sizes in Europe are typically smaller than in the US, meaning that fewer EU firms can access the EUR 100M+ cheques needed for rapid scaling. This was to be seen in the period 2024-2025 when 54% of lead investors in late-stage VC investments in European AI Startups were from American investors. A result of this is that many EU startups either up either accept lower valuations or seek funding from non-European investors, too often resulting in relocation to the US. 

Compounding this issue is the fragmentation of Europe’s capital markets that the EU’s Saving and Investment Union Strategy seeks to address.

Another hurdle is Europe’s volume of spinoff creation and related tech transfer policies, which can prevent groundbreaking research from reaching the market. The EU remains a global leader in R&D, with world-class universities and corporate labs, but it still lags the US in commercialising innovation. Many European universities have historically retained excessive ownership of intellectual property (IP) developed by researchers, discouraging academics from launching spinoffs, although recent years have seen encouraging progress in this area. When spinoffs do emerge, they often lack the funding, mentorship, or business expertise needed to scale.

Progress on tech transfer is systematically measured by the European Startup Network Alliance (ESNA) through its “Existence of Policies for Smooth Tech Transfer” indicator.

While ESNA’s indicator shows Member States progressing in the aggregate, towards a “full” policy framework for spinoffs and tech transfer, there remain several opportunities to improve the conditions for spin-off creation and subsequent market growth. For example, Member States could further enhance conditions by:

·Mobilising government co-funding (grants, loan guarantees, or equity investments) for early-stage spin-offs to reduce risk for private investors.

·Encouraging undertakings of all sizes to invest in spin-offs through tax incentives, government grants, or direct partnerships

·Launching national or regional initiatives to encourage entrepreneurs (for example experienced startup founders) to work more closely with university researchers and contribute their expertise (including but not limited to mentoring of researchers open to entrepreneurial paths).

Another barrier is the difficulties EU startups face in accessing public procurement markets, which represent a EUR 2 trillion annual opportunity. Despite their innovative potential, SMEs and startups win less than 10% of public contracts, often because procurement rules and tender procedures are designed in ways that favour large incumbents. In practice, these processes tend to reward companies with long track records rather than younger and more agile firms.

2.3.Ensuring security for competitive growth

The global cybersecurity landscape is being reshaped by growing geopolitical competition and concentrated digital supply chains, which create systemic vulnerabilities across global markets. Cybersecurity is moving toward AI-enabled attack and defence operations, with threat actors increasingly using automation, generative AI, and large-scale exploitation of cloud - edge infrastructures. Meanwhile, advances in quantum computing require an urgent global transition toward Post-Quantum Cryptography.

.

Similarly, the EU is facing a cybersecurity landscape shaped by the weaponisation of AI, ransomware, growing dependence on untrusted suppliers and a chronic shortage of security expertise. Sophisticated adversaries, including both state and non-state actors, are exploiting vulnerabilities in cross-border ICT infrastructures, automated decision systems and emerging technologies. In addition, there is a clear and urgent need to ramp up efforts on Post Quantum Cryptography transition.

Despite significant progress, Europe remains structurally dependent on non-EU cybersecurity suppliers, and European companies are underrepresented in global cybersecurity leadership. The European market continues to rely predominantly on non-EU industry actors.

According to the Digital Decade Eurobarometer 2026, 91% of Europeans think the EU should cooperate with Member States to reinforce cybersecurity and protection from online threats. In addition, 86% of respondents think the EU should prioritise investments in digital infrastructure and services that are developed and controlled in Europe, and 58% of them would be willing to switch to an EU-based digital service provider even if it means slightly higher costs, pointing to the greater security and reliability as the main motivation for doing so.

In their National Roadmaps, Member States reported 39 measures contributing to increased cybersecurity. Almost half of these measures are dedicated exclusively to cybersecurity, with a total budget of EUR 0.79 billion. The other measures have a broader scope, aiming to support several targets across all areas, with a total budget of EUR 6.8 billion. These initiatives often involve developing national cybersecurity strategies, establishing cybersecurity centres, boosting cybersecurity skills and strengthening cybersecurity capacities in businesses, public services and digital infrastructure. This focus is also reflected in the Member States’ roadmap adjustments.

As geopolitical and economic tensions continue to intensify, cyber threats against the EU have further evolved, with espionage, pre-positioning, and disruptive operations increasingly integrated into state strategies. Cyberespionage targeting EU Member States and EU institutions remains persistent and continuous, with threat actors maintaining long-term access to networks, particularly in government, defence, and critical infrastructure sectors. The convergence between state-sponsored actors and cybercriminal ecosystems has become more pronounced, with states leveraging criminal tools, access brokers, and shared infrastructures to enhance deniability and operational reach.

Within the cyber threat landscape, ransomware remains one of the most impactful threats, but its nature has continued to evolve. Attacks are now predominantly focused on data exfiltration and multi-layered extortion, rather than encryption alone. SMEs are increasingly targeted due to weaker security postures. While law enforcement actions disrupted major groups, the ecosystem has become more fragmented and adaptive, with a proliferation of new ransomware actors and rebranded operations. Critical sectors - including healthcare, transport, and public administration - continue to be heavily affected, with ransomware incidents maintaining a high share of impactful disruptions across the EU.

Supply chain attacks remain a key systemic risk, as attackers exploit dependencies on third-party providers and widely used software components to scale impact across multiple organisations. This risk is compounded by continued reliance on non-EU vendors and complex digital ecosystems, increasing exposure to vulnerabilities and external influence. Threat activity targeting cloud environments, managed services, and open-source software has expanded, reinforcing the potential for cascading effects. Together, these developments confirm a shift from isolated cyber incidents to a more interconnected, systemic, and strategically driven threat landscape. Broader trends also point to growing pressure on cyber resilience, including a persistent cybersecurity skills gap - still estimated in the hundreds of thousands across the EU - and ongoing challenges in public awareness and incident reporting.

At EU level, several laws and initiatives are in place to address some of these challenges, mitigate their impacts, or strengthen the level of cybersecurity across the Union. The proposal for a Cybersecurity Act 2 (CSA2 proposal), adopted on 20 January 2026, clarifies and strengthens the mandate of the European Union Agency for Cybersecurity (ENISA); improves the European Cybersecurity Certification Framework (ECCF) and addresses ICT supply chain security challenges. Specifically, with the trusted ICT supply chain framework, the proposal aims at de-risking the Union’s critical ICT supply chains, starting with the electronic communications sector. Anchored in the ICT supply chain framework, the new ECCF will deliver trust for critical ICT technologies such as 5G and cloud.

The CSA2 proposal also aims to establish a mechanism to validate the skills and experience acquired by cybersecurity professionals against a common set of criteria, defined at European level, and implemented at national level. By developing a mechanism of European individual cybersecurity skills attestations, it will facilitate skills portability and support the single market by supporting the emergence of new providers. The CSA2 proposal will have positive economic impacts by supporting cybersecurity professionals’ mobility, reducing labour and skills shortages in cybersecurity. It will create a market of cybersecurity attestations that are easy to understand for employers and learners, tailored to the European Union labour market.

The NIS2 Directive strengthens the cyber resilience of critical sectors in the EU by requiring essential and important entities in these sectors to take cybersecurity risk-management measures and to report significant incidents. By ensuring organisational measures, such as supply chain security measures, the Directive improves the cybersecurity baseline and supports the continuity of services that are essential for the society. Moreover, the Directive requires Member States put in place national cybersecurity strategies and coordination structures, supporting Member States’ overall cyber preparedness. At European level, the Directive creates cooperation networks (NIS Cooperation Group, CSIRTs Network and EU-CyCLONe), with the objective of promoting trust, confidence and cooperation between Member States. By creating an overarching framework for cybersecurity in critical sectors, the NIS2 Directive also supports the preparation of critical infrastructure sectors against emerging cybersecurity threats, such as cryptographically relevant quantum computers (CRQC).

Concerning the cybersecurity of specific critical sectors, the Commission has continued the implementation of the Action Plan on the cybersecurity of hospitals and healthcare providers, in cooperation with ENISA, the European Cybersecurity Competence Centre (ECCC) and Member States. The ECCC launched a call for proposals, allocating EUR 30 million to reinforce cybersecurity capacities in hospitals and healthcare providers, while ENISA issued guidance on cybersecurity practices in September 2025. ENISA has set up a cybersecurity support centre for hospitals and healthcare providers, financed under a Contribution Agreement through the Digital Europe Programme. Building on these and other relevant actions, the Commission will put forward recommendations to further refine the Action Plan.

As regards the security of products with digital elements, the implementation of the Cyber Resilience Act (CRA) will address insecure hardware and software products circulating in the internal market and being integrated in ICT supply chains across sectors. Products lacking embedded cybersecurity requirements expand the risk surface of SMEs and critical infrastructure as they present exploitable vulnerabilities, creating risks of cascading disruptions and threats to the European economy and society, as showcased by the recent Collins Aerospace incident targeting airline check-in and boarding software. Enforcing mandatory security-by-design and lifecycle obligations will improve Europe’s cyber resilience and make cybersecurity a market differentiator rather than an afterthought. Over time, the new EU cybersecurity regulatory framework will strengthen the EU’s cybersecurity overall preparedness, reduce dependencies on high-risk third-country suppliers and position the EU as a global standard-setter in secure digital ecosystems.

Looking ahead, future cybersecurity investment needs to build on what has already been achieved and focus on areas where strategic gaps remain. Key priorities relate to Europe’s preparedness, digital sovereignty, support regulatory implementation, and enable resilience in the face of accelerating threats and technological change. These priorities can be structured around three strategic investment pillars:

·Pillar 1 - Knowledge & R&I - This pillar strengthens Europe’s technological foundations.

·Pillar 2 - Industrial scale-up & market uptake “Made in the EU” - This pillar consolidates a competitive European cybersecurity industry.

·Pillar 3 - Resilience of the Digital Market and Infrastructures - This pillar strengthens operational resilience.

Cybersecurity must also be treated as a cross-cutting priority. The resilience of digital infrastructures, public services and sectoral systems depends on ensuring that capabilities developed under dedicated cybersecurity programmes are consistently embedded across all investment areas.

3.Protecting and empowering people, reducing burdens and harnessing digitalisation for sustainability

Since its adoption in 2023, the European Declaration on Digital Rights and Principles has informed the Digital Decade Policy Programme, serving as an anchor of the digital transformation of the EU towards its human-centric vision. The principles support the Digital Decade’s targets and guide actions in areas such as skills development, public services, solidarity and digital inclusion.

According to the Digital Decade Eurobarometer 2026, 51% of respondents consider that the EU protects their rights well in the digital environment. In addition, 85% of Europeans think the EU should cooperate with Member States to promote digital education and skills programs, while 84% think it is important that the EU fulfils the objective of ensuring that all EU citizens have basic digital skills (including AI literacy).

3.1.Digital skills for smart society and competitive economy

Empowering citizens and equipping workers with digital skills is at the core of Europe’s digital transformation, in line with the Declaration on Digital Rights and Principles. Digital skills are both a social and an economic imperative. Basic digital skills are essential for meaningful participation in society, access to services, inclusion, and democratic resilience, while the availability of highly skilled professionals, particularly ICT specialists, is critical to Europe’s competitiveness, technological sovereignty, and capacity to deploy advanced digital technologies.

This section therefore addresses both dimensions. It first examines progress and remaining gaps in basic digital skills across the population, with particular attention to unequal access and evolving needs linked to AI and cybersecurity. It then assesses the EU’s capacity to expand its pool of ICT specialists, a prerequisite for innovation, secure digital infrastructure and the wider uptake of advanced technologies across the economy.

In their National Roadmaps, Member States reported investments of EUR 24 billion in basic digital skills (8% of the total budget). The 349 measures reported on the roadmaps primarily focus on improving digital skills in formal education and promoting digital inclusion. This emphasis is also evident in Member States’ roadmap adjustments.

For the training of ICT specialists, Member States reported investments of EUR 11.9 billion (4.1% of the total budget). The 208 measures on ICT specialists mainly focus on increasing the number of people with advanced and highly specialised digital skills, with around one third of these measures targeting individuals in formal education and approximately one quarter focusing on those already in employment. This focus is also reflected in Member States’ roadmap adjustments, which include a sharp increase in measures aimed at boosting advanced digital skills among women.

According to the European Declaration on Digital Rights and Principles, the digital transformation should contribute to a fair and inclusive society and economy that leaves nobody behind. It should benefit everyone, achieve gender balance, and include notably older people, people living in rural areas, persons with disabilities, or marginalised, vulnerable or disenfranchised people and those who act on their behalf. In this regard and according to the 2026 Eurobarometer on the Digital Decade, 90% of Europeans think it is important for the EU to make digital tools more accessible for everyone (especially vulnerable groups, older ones, people with disabilities, etc.). In addition, according to Eurostat , 24.9% of citizens across Europe use the internet for civic or political participation.

3.1.1.Basic digital skills 

Over the last two years, the level of basic digital skills in the EU has increased significantly. The percentage of people with at least a basic level of digital skills rose by 4.8 percentage points, from 55.56% in 2023 to 60.39% in 2025. This represents a year-on-year increase of 4.3%, significantly higher than the yearly growth rate recorded between 2021 and 2023 (1.5%). Some Member States have made significant progress, with year-on-year increases of more than 6%, including Denmark, Romania, Ireland, Italy, Germany, Cyprus and Poland.

However, in 2025, the percentage of people with at least a basic level of digital skills still stood at only 92% of the ideal value of the Digital Decade trajectory (60.39% instead of 65.51%). According to the forecast along the updated baseline trajectory, only 85.6% of the target will be achieved by 2030, implying that only 68.48% of the population is projected to have at least a basic level of digital skills, instead of the target of 80% ( Figure 15 ).

At this pace of progress, it will take eight additional years from the digital decade deadline to reach the full target, forecast to be reached in 2037 if no further actions are taken.

Figure 15: At least basic digital skills in the EU. Historical data, Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

Basic digital skills are essential for meaningful and safe participation in today’s digital economy and society. As shown by the Eurobarometer on Future Needs in Digital Education (2025) 69 , 89% of citizens agree that digital skills are essential for participation in society while more than eight in ten respondents (86%) agree that expanding digital skills training is important for improving the economy of their countries. Yet, a substantial share of Europeans still faces difficulties in acquiring them. Older people, individuals with low educational attainment, persons with disabilities, marginalised groups facing racial or ethnic discrimination, and those who are unemployed or outside the labour market remain particularly at risk of lacking basic digital skills. Disparities are also evident among younger generations: not all young people benefit equally from digital opportunities and factors such as unequal access to digital tools and infrastructure, as well as limited parental support, can significantly shape learning outcomes.

The barriers to acquiring digital skills go well beyond access to devices and internet connectivity. For older people the transition to new technological tools can be intimidating, especially where there is limited prior exposure, low confidence in using digital tools or a lack of age-appropriate training opportunities and guidance. People with lower educational attainment may face a double disadvantage: they may lack both foundational skills and the capacity for self-directed learning needed to improve their digital literacy. For those outside the labour market, digital upskilling may be deprioritised in favour of other skills perceived as more immediately relevant for re-entering employment. Language barriers can further restrict access when learning resources are not available in accessible formats or in learners’ native languages. More broadly, the rapid pace of technological change means that digital skills need to be constantly updated, a task that requires time and resources that many individuals may not have. Under the Union of Skills, Member States are encouraged to strengthen basic skills provided in formal education and expand training opportunities and participation in lifelong learning. The STEM Education Strategic Plan 70 highlights the importance of increasing the talent pipeline in STEM subjects, and among its actions proposes a 5% enrolment target in doctoral programmes in ICT by 2030. The need to address skills shortages in strategic sectors, including ICT and AI, is also reflected in the Council Recommendation on Human Capital 71 . The Commission plans to adopt an education package, which will complement the Action Plan on Basic Skills and the STEM Education Strategic Plan and will aim to establish a robust and inclusive EU digital education ecosystem, which will help Member States support children and young people who struggle with digital skills. Furthermore, the EU-OECD AI literacy framework for primary and secondary education will put forward a common approach to the competences young people need to develop already at school to understand and critically use AI.

Particular attention should be paid to the evolving nature of digital skills needs. The acceleration of AI uptake is changing the skills profile required for full participation in society and the economy. People face increasing concerns over the protection of their sensitive data, as well as the need to securely interact with AI-based tools and services 72 . Education and training systems therefore need to adapt not only to persistent gaps in basic digital skills, but also to new demands related to AI literacy and cybersecurity awareness. This challenge is especially acute for disadvantaged groups, including people with low socio-economic backgrounds, those living in remote areas, older people, persons with disabilities, and adults outside the labour market, who often need more accessible, targeted and flexible learning pathways. It also requires strengthening the capacity of education and training systems, including teachers, trainers and local providers, to respond to changing skills needs in an inclusive and effective way. 

3.1.2.ICT Specialists

The total number of ICT specialists in employment rose by 260 000, from 10.2 million in 2024 to around 10.5 million in 2025. This reflects a 2.6% year-on-year growth in the number of employed ICT specialists. In 2025, ICT specialists still accounted for just 5.0% of total EU employment, substantially stable since the previous year and far from the EU target of 10% of total employment by 2030.

According to the baseline trajectory, around 12.2 million ICT specialists are expected to be employed by 2030, meaning that only 61% of the target will be achieved by 2030 ( Figure 16 ). In 2025, the value reached by this KPI stood at around 78% of the ideal trajectory value, with around 10.5 instead of 13.4 million. The full target - 20 million ICT specialists in employment by 2030 - is forecast to be reached only in 2052 if no further actions are taken.

Figure 16: ICT specialists in the EU. Historical data, Digital Decade (DD) trajectory and revised baseline trajectory towards 2030.

In 2025, women accounted for 19.5% of employed ICT specialists, a figure unchanged since 2024. Over the past decade, the gender gap in ICT employment has remained pronounced, with men consistently outnumbering women by around 60 percentage points-women’s representation fluctuated between 16.2% and 19.5%, while men’s ranged from 80.5% to 83.8% ( Figure 17 ).

Despite a brief decline between 2013 and 2015, the share of women in ICT roles has grown gradually since 2012, albeit at a slow pace ( Figure 17 , left). In contrast, the proportion of men peaked in 2014-2015 and has since declined, yet it still exceeds women’s representation by more than fourfold ( Figure 17 , right).

Figure 17: Percentage of individuals employed in ICT specialists’ occupations in the EU by sex (2012-2025). The solid line shows the time trend since 2012. Left-hand side: percentage of women in total ICT specialists; right-hand side: percentage of men in total ICT specialists. Please note that the range of the y-axis is different in the two charts.

The need for ICT specialists is rising sharply. As the architects of Europe’s digital future, these professionals play a central role in deploying advanced technologies, increasing productivity, and delivering secure and efficient digital services.

The analysis of Online Job Advertisements (OJA) 73 reveals rapid shifts in specific skill requirements. Demand remains strongest for Data Science and Cybersecurity profiles, while cloud security skills are also rising significantly. The successful deployment of AI requires a complex mix of expertise in data management, cybersecurity, systems engineering and software development.

The rapid uptake of digital technologies across sectors, combined with replacement demand linked to retirement will put further pressure on the ICT workforce in the coming years. Shortages and mismatches are unevenly distributed across Member States, regions, sectors and firms, with stronger digital ecosystems generally better placed to attract and retain talent. Combined with the persistently low share of women among ICT specialists, this means that the European economy is still not making full use of its available talent base.

A number of structural challenges help explain this gap:

·There is a considerable mismatch of ICT skills in the European labour market. Many companies report difficulties in finding people with the required profile, a challenge made harder by rapidly changing skills demands linked to AI, the evolving data economy, and growing cybersecurity needs.

·There are key drop-off points (‘leaky pipeline’) for girls and women in ICT fields notably between secondary STEM education and specialised ICT studies, and between ICT qualifications and employment. This suggests that increasing girls’ participation in STEM alone will not be sufficient to close the gender gap. Evidence points to persistent barriers for women in the ICT workplace (‘leaky pipeline after entry’), including gender bias in recruitment and promotion, unequal recognition and pay, limited access to leadership roles, and greater caregiving responsibilities leading to more frequent career breaks.

·The relatively low number of new ICT graduates (tertiary and initial VET) entering the labour market presents a major challenge to reach the Digital Decade target. In 2023, the EU produced only 2.7 tertiary ICT graduates per 1 000 young people, compared to 3.7 in the US and 4.6 in the UK. Although Member States are seeking to adapt Higher Education Programmes to changing skills demands and increase the attractiveness of ICT careers, current efforts remain insufficient.

·Another challenge lies in the upskilling of existing ICT specialists. Businesses must ensure that their employees remain up to date with the latest ICT skills requirements, which requires substantial investment. In addition, a growing number of sector specialists such as healthcare professionals using AI-powered screening tools, need advanced digital skills to use digital technologies effectively. This increases competition for talent and calls for more dedicated approaches to ICT education, enabling workers with non-technical backgrounds to interact efficiently and safely with advanced digital technologies.

·The concentration of the ICT ecosystem around individual vendors also presents considerable challenges for upskilling and reskilling. As technologies become increasingly vendor-specific, professionals are often required to obtain specialised skills and certifications tailored to particular applications or platforms. Such dependencies may reduce the portability of skills across employers and technologies, make training pathways more fragmented and costly and weaken workforce adaptability and resilience.

Addressing these challenges will require stronger investment in the development and updating of education and training provision, particularly in cooperation with Industry, as well as more flexible learning pathways and netter translation of research outcomes into practical training content. It will also require stronger efforts towards attract and retain women in ICT studies and careers, more support for workforce and in-company upskilling and greater attention to the territorial and sectoral distribution of ICT talent, so that regions and smaller firms are not left behind.

3.1.3.Protecting people, in particular minors, in the online space

Protecting citizens, in particular minors, from harmful and misleading content online remains one of the most pressing challenges of the digital age. In 2025, 55.9% of EU citizens aged 16-74 declared having been exposed to untrue or doubtful content online, an increase from 49.25% since 2023. This corresponds to an annual growth rate of 6.5 percentage points. This trend is also observed in the younger population, with 61.66% of individuals aged 16 to 24 exposed in 2023, increasing to 66.34% in 2025, reflecting an annual growth rate of 3.7 percentage points. There is a notable gap favouring young people, with a difference of 7.77 percentage points between those aged 16-24 and those aged 25-64. For the latter group, the exposure increased from 51.7% in 2023 to 58.57% in 2025, with an annual growth rate of 6.4 percentage points.

Turning to online verification of information, the European Union saw an increase from 24.29% in 2023 to 29.16% in 2025 for all individuals aged 16-74, indicating an annual growth rate of 9.6 percentage points. Among individuals aged 16 to 24, the percentage increased from 34.68% in 2023 to 39.49% in 2025, with an annual growth rate of 6.7 percentage points. The gap between young people and those aged 25-64 is 9.09 percentage points. For individuals aged 25 to 64, the percentage rose from 25.18% in 2023 to 30.4% in 2025, with an annual growth rate of 9.9 percentage points.

Regarding online exposure to hostile or degrading messages, the European Union experienced an increase from 33.5% in 2023 to 39.72% in 2025 for all individuals aged 16-74, reflecting an annual growth rate of 8.9 percentage points. For those aged 16 to 24, the exposure increased from 47.54% in 2023 to 52.99% in 2025, with an annual growth rate of 5.6 percentage points. The gap between young people and those aged 25-64 is 11.85 percentage points. For individuals aged 25 to 64, the exposure rose from 34.53% in 2023 to 41.14% in 2025, with an annual growth rate of 9.2 percentage points.

In summary, the data from the Eurostat surveys reveals a consistent increase in online exposure to both untrue or doubtful content and hostile or degrading messages among EU citizens from 2023 to 2025, with younger individuals particularly at risk. Encouragingly, this is accompanied by an increase in the practice of information verification, again most pronounced among younger individuals. While these trends suggest a growing awareness of online risks, the overall picture underlines the need for policymakers to enhance digital literacy programmes and strengthen measures to protect EU citizens in the digital space. The significant gaps between age groups should be considered when designing targeted interventions.

Preserving information integrity

With the digitalisation of the information space, citizens increasingly face challenges in accessing a plurality of information from independent and reliable sources. The 2025 Media Pluralism Monitor shows that there is a medium to high risk to media pluralism in 23 EU Member States. Additionally, online platforms and AI services increasingly capture news revenues, putting the news sector under significant economic pressure, ultimately threatening media independence and pluralism. This shows the need to strengthen financing mechanisms for news media, in ways that respect media independence, as well as to support news organisations in better harnessing new technological tools.

To counter the escalating threats of disinformation, foreign interference, and algorithmic amplification of harmful content, evidenced by the 6.5 percentage-point annual increase in exposure to untrue or doubtful content (2023-2025) and the growing fragmentation of media pluralism, Member States must fully utilise the EU’s existing legislative framework and non-binding instruments. These include the Digital Services Act (DSA), the Political Advertising Regulation, the AI Act, and the European Media Freedom Act (EMFA), as well as the voluntary Code of Conduct on Disinformation, which taken together form a cornerstone for systemic resilience. Several measures under the European Democracy Shield (EDS) as well as the EDMO Hubs provide important supportive frameworks, which complement efforts to ensure rigorous, coordinated enforcement of relevant EU legislative frameworks, both at EU and national levels.

Generative AI represents another evolving challenge. According to the latest Eurostat data , there is substantial uptake of generative AI tools, especially by the younger demographic. In 2025, 63.8% of young people aged 16-24 in the EU used generative artificial intelligence (AI) tools compared to just 32.7% of those aged 16-72.  Reports have shown that foreign powers have been trying to poison LLM training data through techniques such as AI grooming (one example of this is Russia’s Pravda network). Together with other problems noted in AI chatbots such as hallucinations or sycophancy, this creates a risk of untrue or doubtful content or even manipulation especially for the younger segment of the population.

Protecting children

According to the 2026 Digital Decade Eurobarometer survey, an overwhelming majority of respondents (93%) think it should be a high priority for the EU to further strengthen the protection of children and young people online.

With growing public concern over the risks children face in digital spaces, it is important to continue enhancing their online safety and well-being through robust regulatory and enforcement measures. In response to the demands by citizens, the Commission has been working on several streams to increase the well-being and safety of children online.

The General Data Protection Regulation (GDPR) already recognises that children merit specific protection regarding the processing of their personal data. Accordingly, organisations must implement child-specific data protection measures to provide a higher level of protection against the risks that may arise from children’s use of digital services. The GDPR also contains specific safeguards for minors, including the ‘right to be forgotten’, and establishes age-of-consent requirements for the processing of personal data by information society services, such as social media, online gaming and other digital platforms.

Further reinforcing these protections, Article 28 of the Digital Services Act (DSA) requires providers of online platforms of all sizes accessible to minors to ensure a high level of privacy, safety and security for minors on their services. To further boost online safety for children and young people under the DSA, the Commission adopted the Guidelines on the protection of minors in 2025. These guidelines set out a benchmark that the Commission uses to determine compliance in this area and outlines a non-exhaustive list of proportionate and appropriate measures to protect children from online risks such as grooming, harmful content, problematic and addictive behaviours, as well as cyberbullying and harmful commercial practices that may occur on online platforms accessible to minors. First enforcement actions have been taken against the providers of TikTok, Facebook and Instagram as well as pornographic-content online platforms. These investigations concern issues such as accessing age-inappropriate content, addictive behaviour, account settings and appropriate age assurance methods. Additionally, Digital Services Coordinators (DSCs) have started to enforce Article 28 of the DSA at the national level 74 .

To strengthen protection for young audiences in an evolving digital landscape, the Audiovisual Media Services Directive (AVMSD) imposes obligations on video sharing platforms providers to embed EU content standards and child-safeguarding measures into their operations. Under the AVMSD, all video-sharing platforms (VSPs) must include EU media content standards, notably those aimed at protecting minors from harmful content, in their terms and conditions. Their providers are also required to take appropriate measures (e.g. mechanisms to report or flag harmful content, age verification, parental control and content rating systems) to prevent minors from viewing harmful content. The Commission continues to monitor the implementation of the AVMSD by the Member States, including the provisions pertaining to the protection of minors. In addition, the Directive’s ongoing evaluation and review planned by Q3 2026 will assess whether more should be done to ensure that minors are protected when they view audiovisual content online, including when made available by influencers when they qualify as audiovisual media service providers.

In response to the growing epidemic of cyberbullying 75  the Commission has adopted an action plan against cyberbullying 76 . The initiative aims to strengthen the capacity to prevent, report and combat cyberbullying. The plan was prepared through targeted and public consultations involving children, researchers, experts, and the wider public.

The action plan is structured around three pillars:

·A coordinated EU approach to protection: The Commission will enforce existing laws while strengthening their focus on cyberbullying and invites Member States to develop national policies based on a shared understanding of the issue.

·Prevention and awareness: The plan promotes responsible digital habits from an early age and will provide EU-level tools developed with input from key stakeholders.

·Reporting and support: The plan ensures clear and accessible reporting and support for all, especially victims and bystanders. The Commission will support the rollout of an online safety app across Member States to help children report cyberbullying, store evidence and access assistance.

Lastly, efforts to establish an EU-wide approach to age verification are advancing, with the release of the blueprint for a secure, privacy-preserving and fully data-protection-compliant EU Age Verification solution in July 2025. On 15 April 2026, the Commission presented the feature-complete EU age verification solution. In parallel, and to accelerate progress across the EU, the Commission adopted a recommendation on 29 April 2026 urging Member States to make use of the age verification blueprint and draw up implementation plans to ensure swift adoption of national age verification solutions by 31 December 2026. The forerunner Member States (DK, FR, GR, IT, ES, CY and IE) are advancing with their implementation of the Age Verification solution, and the first national solutions are expected to be available mid-2026.

3.2.Efficient public services and administrative burden reduction

This section examines how the digital transformation of public services can improve efficiency, reduce administrative burden, and enhance accessibility for citizens and businesses across the Union. Building on the Digital Decade principles, it focuses on key enablers such as secure digital identity, interoperable services, and access to essential public services online. While progress is ongoing, further efforts are needed to streamline procedures, strengthen cross-border functionality, and ensure that digital solutions deliver tangible simplification benefits.

The European Digital Rights framework requires all Member States to offer citizens an accessible, voluntary, secure and trusted digital identity. This should in turn allow people to access a range of online services, including medical records and other healthcare data. According to the Digital Decade Eurobarometer 2026, 79% of Europeans think EU should cooperate with Member States to develop shared digital public services (e.g. digital ID, e-Health).

3.2.1.European Digital Identity and business wallets

The European Digital Identity (EUDI) Framework is a key enabler of Digital Decade targets. Electronic identification 77 allows people to securely verify their identity and access services across the EU. According to 2025 data, 52% of people aged 16-74 in the EU stated that they had used their eID to access online services for private purposes in the previous 12 months. Results vary significantly across Member States, from over 90% in Denmark (99%), Finland (96%), the Netherlands (95%), Sweden (92%), and Estonia (91%) to below 15% in Germany (15%), Slovakia (14%) and Bulgaria (12%). Some countries are showing rapid progress, with Cyprus improving its score from above 9% (2024) to 57% (2025).