Lucia Russo
Jeff Mollins
Bénédicte Rispal
Peter Wittemann
OECD Economic Surveys: Slovak Republic 2026
Abstract
Artificial intelligence can enhance innovation and productivity across the Slovak economy. While AI adoption is growing, it remains below EU averages, especially among SMEs. The chapter highlights the need to expand access to capital, data, and computing infrastructure, and to strengthen advanced research and public-private collaboration. It also addresses regulatory challenges to implement the EU AI Act and emphasises the role of regulatory sandboxes to support experimentation. Finally, it underscores the urgency of tackling AI-related skills shortages through education reform, workforce training and talent retention efforts.
Artificial Intelligence (AI) has emerged as a transformative force, offering the potential to significantly enhance productivity, stimulate economic growth, alleviate skills gaps and help address structural challenges (see Chapter 2 of this Economic Survey). An OECD study estimates potential gains for Slovakia from AI in the range of 0.2 to 0.8 percentage points in annual labour productivity growth over the next decade (Filippucci et al., 2026[1]). These estimates come with considerable uncertainty that surrounds the economic effects of current AI technologies, forthcoming advances in AI and, crucially, the speed and depth of AI adoption across the Slovak economy. In this context, AI shows promise as a general-purpose technology (Calvino, Haerle and Liu, 2025[2]), but productivity gains will hinge on ecosystems that enable widespread and effective adoption.
The 2024 figures of the OECD.AI Index (OECD, 2026[3]) - which provide a holistic view of national AI ecosystems using a composite measure - place Slovakia relatively low compared to OECD counterparts, including Hungary, Poland and Czechia (Figure 3.1). The country scores particularly low in both the R&D and policy environment components, highlighting gaps with other OECD members in advancing AI research and in fostering an ecosystem capable of bringing innovations to the market, both in terms of investment capacity and regulatory agility.
The AI Vision for Slovakia unveiled in December 2025 sets out the aims of strengthening digital sovereignty, accelerating innovation and deploying AI widely. The vision points to the country’s low-carbon energy mix (nuclear and hydro, see Chapter 4) as an asset for the development of AI computing and data infrastructure in line with EU initiatives such as AI Factories. For the country to realise the positive gains from AI across the economy, it must achieve broader adoption, particularly among SMEs, by supporting companies’ investments in complementary assets and workers’ digital skills. It is important to build AI-related competences by incorporating them in the education system as set out in the 2025 Plan for Responsible Use of AI in Education.
Potential gains are large in Slovak industry including the automotive, pharmaceutical and machinery sectors. AI adoption can also enhance productivity in services ranging from government to healthcare, financial and tourism services through improved delivery and more data-driven decision-making. More broadly, integrating AI in these sectors and beyond can help make growth more knowledge-based and innovation-driven (see also Chapter 2).
Index scores, from 0 to 1 (highest performance)
Note: Scores are based on a composite relative index with 1 being the highest possible value.
Source: (OECD, 2026[3]).
Uptake of AI is still at a relatively early stage Slovakia. In 2025, 18% of enterprises reported using AI, a marked increase from 7% in 2023. This adoption rate, while it is comparable to regional peers such as Czechia and notably higher than Hungary and Poland’s, remains below the EU average across company sizes (Figure 3.2).
Like with other technologies, company size is a key determinant of AI adoption. As in other OECD and EU countries, larger companies in Slovakia are considerably more likely to adopt AI than small businesses, with adoption rates three times higher. This reflects differences in investment capacity, access to complementary assets, organisational readiness, and digital maturity (Calvino and Fontanelli, 2023[4]). However, only 29% of large Slovak enterprises use AI, compared to 41% in the EU. Among medium-sized companies, adoption stands at 16% (EU: 21%), and for small businesses, just 9% (EU: 11%) (Figure 3.2). This suggests ample scope for Slovak companies, including large enterprises, to accelerate AI adoption.
% of enterprises, 2025
Note: AI adoption rates across countries small (10-49 employees), medium (50-249 employees), and large (250 or more employees) enterprises.
Source: Eurostat (2025[5]), Statistics Explained: Use of Artificial Intelligence in Enterprises, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Use_of_artificial_intelligence_in_enterprises, accessed ion 19 December 2025.
Patterns of AI adoption vary widely across sectors, reflecting differences in digital intensity, innovation capacity, technologies, and company demographics, including workforce skills. As in other EU countries, digital-intensive industries in Slovakia lead in AI use, but adoption levels are lower than the EU average across nearly all sectors. In the ICT sector, 50% of Slovak enterprises used AI in 2025, below the EU average (63%) and Czechia (60%) but above Hungary (46%) and Poland (38%). The professional, scientific and technical services sector (28% in 2025) is also below the EU average (40%) and behind Czechia (33%), although above Hungary (23%) and Poland (22%).
By contrast, adoption has progressed well in administrative and support service activities at 20% in 2025 in line with the EU average. This has been driven by the outsourcing of corporate support functions (shared-service and business process centres) by multinational enterprises, with 80 service centres, including those of AT&T and IBM, established over the past two decades (SARIO, 2025[6]). Likewise, Slovakia exceeds average EU adoption rates of AI in the transportation and storage sector, reflecting the logistics base and integration in regional supply chains.
AI can help develop core manufacturing functions. For instance, AI can help detect anomalies in machinery and forecast equipment failure in predictive maintenance and estimate future energy needs to optimise production schedules (OECD, 2022[7]). AI can also help optimise supply chain operations, such as inventory management and routing decisions, and improve processes and job scheduling to reduce costs and enhance productivity (Milanez, 2023[8]). In addition, AI contributes to quality assurance by detecting visual defects in products, to workplace safety by identifying hazards, and to robotics and automation by enabling object recognition, path planning, and human-machine interaction (OECD, 2026[9]). In the automotive sector, AI can provide crucial support in the shift from internal combustion engine vehicles to electric vehicles (EVs) – by optimising EV production, accelerating battery innovation, and helping to manage energy consumption – as well as in supporting automated driving and security-enhancing functionalities.
In 2024, 8% of manufacturing companies in Slovakia reported using AI, below the EU average (11%) and Czechia (10%), although ahead of Hungary (5%) and Poland (5%). Within manufacturing, adoption differs sharply by subsector (Figure 3.3). In 2024, 18% of companies in machinery and equipment and 16% in the automotive sector reported using AI, both above EU averages (15%) (Figure 3.3). These subsectors have benefited from historically large foreign direct investment, have many large companies and are well integrated in global value chains, facilitating the adoption of advanced technologies. By contrast, AI use in more traditional and lower value-added industries (food, beverages and tobacco, wood and paper, textiles and apparel, rubber and plastics, and basic metals) remains limited as well as in the pharmaceutical sector (Figure 3.3).
% of enterprises with 10 employees or more adopting AI, by sector, 2024
Source:Eurostat (2025[5]), Statistics Explained: Use of Artificial Intelligence in Enterprises, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Use_of_artificial_intelligence_in_enterprises, accessed in August 2025.
Slovak automotive manufacturing companies perform relatively well in logistics AI (25% of companies that used AI in 2024 reported this function), ahead of the EU27 average (19%) and comparable to regional peers. However, fewer than one third of enterprises using AI in 2024 reported deploying AI in their production processes (28%), significantly behind Poland (64%) and Hungary (77%). Most concerning is the very low use of AI in R&D and innovation (10%), which is half the EU average (19%) and far below Poland (18%). These statistics reflect Slovakia’s position as a production hub whereas innovative activities by multinational carmakers are taking place abroad (see Chapter 2).
Uptake of automation-oriented AI such as robotic process automation and autonomous robots is comparably high in Slovakia, especially in the automotive and machinery sectors. This is in line with the robot density in the sector, which was 201 per 10,000 in the automotive sector in 2023, above the world average although still below the EU level (International Federation of Robotics, 2024[10]). This strong automation base is largely driven by foreign carmakers operating in Slovakia, such as Volkswagen, Stellantis-PSA, Kia, and Jaguar Land Rover.
The low figures of AI use in automotive production processes and the differences in AI technologies used by company size (Figure 3.4) suggest a two-speed system whereby global Original Equipment Manufacturers (OEMs) make use of AI in production processes, while Slovak enterprises, in particular SMEs, have yet to integrate AI into their activities. National initiatives such as the “National Project for the Introduction of Industry 4.0 elements and digital transformation” play a central role by partnering with universities (TUKE, UNIZA, STUBA) and supporting industrial enterprises through consulting, digital readiness audits, transformation planning, and implementation. The strategic framework of the Research and Innovation Strategy for Smart Specialisation (RIS3 SK), especially its updated “Innovative Industry for the 21st Century” domain, aims to further reinforce the integration of digitalisation and AI within industrial production. Swift implementation of the framework targets, particularly increasing thematic support for automation and robotisation as well as fostering collaboration across private and academic sectors, is key to advance AI uptake within the automotive industry while ensuring knowledge spillovers to local research institutions and the rest of the economy.
3.3.1. There is scope for broader AI use across companies
AI used in Slovak business processes is concentrated in activities such as marketing and production. Uptake in more knowledge-intensive functions, including research and development (R&D) and innovation, is comparatively limited (Figure 3.4). Larger companies are significantly more likely to employ AI across multiple functions, whereas SMEs mostly use AI for marketing and sales, in line with OECD trends and increasing use of generative AI by SMEs (Kergroach and Héritier, 2025[11]; OECD, 2025[12]; OECD, 2025[13]).
% of enterprises using at least one AI technology, 2024
Note: AI adoption rates across countries small (10-49 employees), medium (50-249 employees), and large (250 or more employees) enterprises.
Source:Eurostat (2025[5]), Statistics Explained: Use of Artificial Intelligence in Enterprises, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Use_of_artificial_intelligence_in_enterprises, accessed in August 2025.
The distribution of AI use also varies markedly by technology. In line with patterns at EU level, large companies report higher adoption rates of advanced AI tools such as machine learning, natural language processing and computer vision. In contrast, SMEs tend to rely on more accessible applications, including chatbots. AI-based robotics and image recognition – key for AI-enabled quality control in manufacturing – in particular, are concentrated among large companies, raising concerns that SMEs may not be able to fully exploit the productivity-enhancing potential of AI.
To support further adoption of digital technologies, including AI, Slovakia has established five European Digital Innovation Hubs (EDIHs) (European Commission, 2025[14]), which serve as one-stop shops for advancing digital transformation in businesses and the public sector. These hubs provide resources, expertise, and support for companies, with one of them - EXPANDI 4.0 - promoting broad industry digitalisation including AI, IoT, big data, Augmented and Virtual Reality, predictive maintenance, and robotics. According to the latest EU-wide EDIH report (European Commission: Joint Research Centre, 2025), Slovak hubs rank 17th in customer reach per capita, behind Czechia and Hungary, suggesting room for broader engagement. At the same time, over 20% of services are delivered outside home regions, showing flexibility and potential to scale impact nationally. Support through EDIHs should be strengthened to increase AI uptake and data readiness of companies, particularly SMEs.
3.3.2. AI is being rolled out across the public sector
AI-driven tools, including chatbots and algorithmic systems, are also being adopted in Slovakia’s public sector. Current applications include AI chatbots on platforms such as slovensko.sk to handle citizen queries, predictive analytics in social services to identify vulnerable groups, and RPA for routine administrative tasks like business registrations.
Looking ahead, the National Concept for the Informatization/Digitalisation of Public Administration (MIRRI, 2021[15]) outlines a plan for more advanced uses across public administration, ranging from health and social services to education and justice. The 2021 National Concept, together with the AI Vision for Slovakia adopted in December 2025, highlight the importance of improving data quality and interoperability across public registers, strengthening central coordination for AI and data governance, and addressing long-standing institutional fragmentation within public administration. The National Concept, as the core framework for digital transformation of public administration to 2030, foresees the systematic integration of AI into government processes, including AI assistants for civil servants, AI-supported decision-making, interoperability standards, and auditability mechanisms for AI systems. Examples include life-event management, and algorithmic decision-making in welfare and minor judicial cases. Plans also include deploying fraud detection algorithms in procurement. Further inspiration can be drawn from international experiences with AI across public administration (Box 3.1).
A key area where AI can fundamentally modernise public service delivery is education. The Plan for Responsible AI Use in education identifies significant potential for AI deployment in the sector starting with AI-supported personalised learning, early-warning systems for dropouts and administrative automation (Ministry of Education, Research, Development and Youth, 2025[16]). The plan, which builds on two digital transformation initiatives, DITEDU and DigiEDU, foresees teacher training together with the deployment of digital resources and infrastructure across educational institutions. The AI Integration Plan aims to alleviate the workload of teachers by automating routine tasks, allowing them to allocate more time to students and tailor their lessons more effectively, also using AI (Ministry of Education, Research, Development and Youth, 2025[17]).
Common challenges that governments face in adopting and scaling AI include skills shortages, limited access to high-quality data, outdated digital systems, funding constraints, and distinct regulatory requirements among government functions (OECD, 2025[18]).The Slovak government has introduced AI training for civil servants (see section on skills), while there are plans to develop a central function to improve cross-agency interoperability. Public authorities should ensure effective implementation of these initiatives, while strengthening inter-ministerial coordination to facilitate the exchange of best practices and foster scalable and trustworthy AI solutions across the public sector.
Several countries have assigned responsibilities to a central dedicated institution to accelerate AI deployment across public bodies while ensuring consistency. Examples include Norway with the Ministry of Digitalisation and Public Governance, Spain with the State Secretariat for Digitalisation and AI and the United Kingdom with the Department for Science, Innovation and Technology (OECD, 2025[18]). Denmark and Sweden have put in place initiatives to foster more coordinated approaches to AI across lower levels of government (OECD, 2025[18])
In the healthcare sector, Slovakia has made efforts to build a strong health data governance system since 2005, centered on the National Health Information System (NCZI) and mandatory electronic health records (EHRs). The eHealth (eZdravie) system, launched in 2018, enables secure data sharing among healthcare providers, and supports the integration of AI in diagnostic and administrative processes. Centrally managed by the NCZI, the system promotes interoperability through international standards and underpins AI integration, while new frameworks for the secondary use of health data aim to support research and innovation. Nevertheless, uneven adoption among healthcare professionals, limited institutional coordination, and concerns over the NCZI’s capacity to manage increasing data volumes remain ongoing challenges. Under its Digital Transformation Strategy to 2030, advance personalised mobile medicine, improve EHR and medical imaging sharing, and strengthen safeguards for sensitive health data (OECD, 2025[19]) Addressing governance and coordination gaps will be essential to realising these objectives effectively.
Education is another sector facing data-related regulatory challenges when deploying AI. The implementation of the EU AI Act is particularly affecting high schools and VET institutions planning to use AI-based educational platforms. To comply with the AI act without constraining pedagogical innovation, educational institutions need guidance on how to protect data, articulate AI tools and human oversight, and ensure transparency.
Public administrations across the world are rapidly deploying a vast array of AI-based initiatives. The five examples below, selected from over 200 reported in OECD (2025[18]), offer a glimpse into the breadth of potential areas where AI can support government activity:
AI for estimating compliance costs (Germany)
Germany’s Federal Statistical Office is developing a machine learning tool to support regulatory impact assessments by estimating compliance costs. The system identifies relevant legal text, predicts cost implications of proposed changes, and automates low-risk estimations, while complex cases remain under human review.
AI for judicial document management (Spain)
Spain’s Ministry of Justice built AI tools to classify, summarise and anonymise judicial documents. Integrated with the national e-justice platform, the system automates data extraction and pre-fills forms, cutting processing time from minutes to seconds and improving overall justice system efficiency.
AI-powered citizen engagement (Finland)
The City of Helsinki deployed UrbanistAI, a tool that generates visualisations of alternative urban planning scenarios. This enables citizens to better understand proposals and participate in decision-making, fostering transparency and consensus in complex planning processes.
AI for procurement oversight (Brazil)
Brazil’s federal audit authority developed an AI system that analyses bids, contracts, and public notices to detect fraud and inefficiencies. It uses automated text analysis and process automation to monitor procurement in real time, flag risks, and accelerate audits, cutting processing time from 400 days to 8 and delivering substantial financial savings.
Sources: Governing with Artificial Intelligence (OECD, 2025[18]), Public Sector Tech Watch (European Comission, 2025[20]).
Slovakia is lagging on advanced research in AI. This is apparent in a low level of national research output, primarily by academic institutions. While AI-related research has increased steadily in Slovakia in recent years (Figure 3.5), output remains low relative to OECD averages (Figure 3.6). Since 2018 and the establishment in 2019 of the Slovak Centre for Artificial Intelligence Research, which serves as a national platform of excellence, AI-related scientific fractional publications have more than tripled, with citations rising at a similar pace (OECD.AI, 2025[21]). Nonetheless, Slovakia continues to lag most OECD countries in terms of research output. In 2024, Slovak research accounted for only 0.1% of global AI-related publications, equivalent to approximately 46 publications per million people (Figure 3.6). This compares to an OECD average of over twice that at about 92 publications per million people. Citation data tell a similar story: Slovak publications have had about 2,000 fractional citations in 2022, also representing about 0.1% of global total citations on AI related publications.
Developing an AI ecosystem that brings skilled workers and necessary computational resources is essential to facilitate innovation. Policymakers have taken steps to strengthen Slovakia’s AI ecosystem. The country has established European Digital Innovation Hubs (EDIH) and the Slovak AI Digital Innovation Hub (SKAI‑eDIH). The Kempelen Institute of Intelligent Technologies (KInIT) -- an independent, non-profit research institute dedicated to intelligent technology research – serves as a coordinator of the EDIH. KInIT brings together and nurtures experts in AI and other areas of computer science, from both private and public sectors. This institute has begun the lorAI (Low-Resource Artificial Intelligence) project, which focuses on developing AI technologies suited for environments with limited data or computational resources. This project targets areas such as natural language processing in under-resourced languages (including Slovak), fact-checking and energy-efficient AI models. This project will be particularly well-positioned to support Slovak use cases given the low quantity of open datasets in the local language. Acting on the objective in the RIS3 SK+ strategy to intensify support for AI research, additional steps could be taken to support the development of an AI ecosystem, including strengthening the collaboration between the public sector and academia through research grants.
Total fractional AI-related scientific publications
Note: Publications are expressed as fractional counts, where equal weights are assigned to each publication’s co-author to avoid double-counting of publications.
Source: OECD.AI (2025), data from Elsevier. Please see https://oecd.ai/en/elsevier for more details.
Fractional publications per million people, 2024
Note: Publications are expressed as fractional counts, where equal weights are assigned to each publication’s co-author to avoid double-counting of publications.
Source: OECD.AI (2025), data from Elsevier. Please see https://oecd.ai/en/elsevier for more details.
Difficult access to venture capital (VC) complicates AI development. Companies in Slovakia receive relatively low levels of VC funding (see Chapter 4 of this Economic Survey). In 2024, total VC investment in Slovak companies reached USD 156 million across eight deals. AI-related companies attracted three of those deals but had a combined value of only USD 8.7 million. This was a sharp decline from USD 35.8 million in 2023. On a per capita basis, this represents just USD 1.6 per person and 0.55 VC investment deals per million people, placing Slovakia among the lowest in the OECD. By comparison, the OECD average is about 41 USD per person and 3.7 deals per million people. Over the past few years, Slovakian AI companies attracting VC funding have reflected a variety of industries, including healthcare, robots and IT hardware, and the energy and mining sector (OECD.AI, 2025[22]).
Studies also point to relatively low AI investment in Slovakia. A recent report by the OECD (OECD, 2025[23]) estimates Slovak combined public and private AI-related investment expenditures to be about EUR 15 per capita, well below the EU27 average of EUR 55 per capita. However, these estimates rely on patent application data to calculate the intensity of AI in total R&D spending. Patenting activity is well below the OECD average in Slovakia, which may reflect low innovation or a preference to avoid patenting when possible. If such a preference contributed to low patenting rates, it would bias the AI intensity estimate downwards. Nevertheless, robustness checks using alternative ways to identify AI spending confirm that per capita public R&D in AI is among the lowest in the European Union (OECD, 2025[23]).
High upfront investment costs and overall financial demands remain key obstacles for companies investing in AI, including those in the manufacturing sector (Milanez, 2023[8]). Although funding mechanisms exist, they are often seen as bureaucratic and not well-suited to the fast-paced nature of AI development. Slovakia is currently addressing this gap partly through the provision of digital and innovation vouchers, alongside other grant programmes under its Recovery and Resilience Plan. The voucher schemes, with a combined budget of over EUR 23 million, typically provide SMEs with support ranging from EUR 1 000 to EUR 15 000 per project to foster digitalisation and R&D collaboration with research institutions. While not exclusively AI-specific, AI-related projects are eligible. An evaluation of voucher schemes in Slovakia highlights their effectiveness in stimulating private R&D spending and fostering collaboration (VAIA, 2025[24]). However, the evaluation also recommended to further simplify administrative processes and broaden the scope of support to achieve systemic impact.
Another obstacle for AI adoption and innovation in Slovakia is the relatively low availability of high-quality data. The Strategy for the Digital Transformation of Slovakia 2030 (2022[25]), which emphasises AI, identifies insufficient high-quality data from public as well as private sources as a major bottleneck. Availability of and access to high-quality data are increasingly recognised as foundational enablers of AI development. Without access to relevant, representative, and well-curated datasets - particularly in the local language - researchers and developers are unable to train, validate, or deploy models effectively. An open data portal has been established through the 2030 Digital Transformation Strategy for Slovakia. A recent legislative proposal seeks to expand access to high-value datasets by enabling their secure re-use for research and innovation while stopping short of making them fully open. An option to further facilitate AI development would be to allow open access to securely anonymised versions of the datasets (OECD, 2025[26]).
One exogeneous hurdle for local large-language-model development is the limited number of Slovak speakers by comparison with the most widely used languages. By this measure, Slovak ranks 149th. Nevertheless, Slovak is represented in 462 open training datasets, the 40th rank among all languages, according to Hugging Face data as of October 2025. Despite this being elevated relative to Slovak’s rank by number of speakers, this is a significantly lower number than for the top languages of English (34,924 datasets), French (5,205 datasets), and Chinese (3,300 datasets). Slovak has only 32 dedicated monolingual datasets available, ranking 66th worldwide (Figure 3.7). This situation underlines the importance of facilitating access to existing Slovak-language datasets.
Number of open monolingual training datasets by language, logarithmic scale
Source OECD.AI (2025), using data from Hugging Face available at: https://oecd.ai/en/data?selectedArea=ai-models-and-datasets.
Another limiting factor is scarce physical infrastructure, also a consequence of narrow digitalisation (see Chapter 2 for more discussion of the digital readiness). Slovakia hosts just 13 data centrer, about 2.4 per million people, below the OECD average of 8 data centres per million people (Figure 3.8). A relatively narrow base of domestic data centres increases the latency of data access, weighing on performance in AI development. Access to the transmission grid is a crucial requirement for data centres or large-scale computing platforms, so-called “hyperscalers” (Cremona and Czyzak, 2025[27]). For this purpose, as well as to ease the development of low-carbon power-generation sources, it is essential to accelerate investment in the power grid as foreseen in the National Energy and Climate Plan for 2021-2030 (see also Chapter 4). Cutting red tape and more broadly reforming regulations to improve the business environment (see Chapter 4) would also facilitate the installation of data centres. By contrast, there has been substantial progress with the laying out of fiber optic cables, ideal for AI workloads due to their low latency and high bandwidth capabilities, which now represent almost 50 percent of total fixed broadband subscriptions (see Chapter 4).
Data and infrastructure constraints likely contribute to the observed absence of locally developed AI models according to the AIKoD and Epoch databases (Epoch AI, 2025[28]; OECD.AI, 2025[29]). Despite these findings, there are some positive signs of AI development in the country. For example, about 1.82 percent of enterprises with 10 or more employees in Slovakia are using AI technologies developed by their own employees in 2024 (Eurostat, 2025[5]). While this remains below the average of the 23 European countries surveyed, it is nearly double the percentage reported for Slovakia in 2021. Expressed differently, 17 percent of enterprises that are using at least one type of AI technology are using AI developed internally. As such, progress in AI adoption could lead to adaptation of AI technologies to the needs at hand, ultimately encouraging the development of local models.
Data centres per million people
Source: Data Centre Map (2025[30]) and OECD calculations.
Ongoing advances in AI rely on large-scale models that demand vast computational resources (OECD, 2023[31]). At present, Slovakia lacks domestically available high-performance computing infrastructure, such as supercomputers (TOP500, 2025[32]), limiting domestic AI research capacity. While IT systems are globally integrated, increasing local computing capacity would still be beneficial. For developers of AI models, purchasing compute resources from abroad, including through cloud providers, remains an option, but it is an imperfect substitute for local infrastructure due to the need to transfer large volumes of data for processing. Several initiatives in Slovakia aim to close this gap. A powerful new supercomputer, named Perun, was put into service at the Technical University of Kosice on 20 November 2025. Perun is positioned among the top 10 most energy-efficient supercomputers globally. It will support research in AI, natural sciences, and engineering, aligning with Slovakia’s goal to enhance its European high-performance computing capacities. Its modular design also provides scalability to meet future demand.
As an EU Member State, the Slovak Republic is preparing to implement the European Union’s AI Act (Regulation EU 2024/1689), which establishes a risk-based framework classifying AI systems into tiers (minimal, limited, high, and unacceptable risk), each with corresponding regulatory obligations. High-risk AI systems must meet strict safety, transparency, and quality requirements and undergo conformity assessments before deployment. Compliance with these rules will require companies to adapt their processes and infrastructure. In practice, developing a high-risk AI system means implementing controls such as a quality management system, technical documentation, risk assessments, and human oversight mechanisms to ensure the AI’s safety. The Slovak Republic is advancing its national AI legislation, which aims to align with the EU AI Act while establishing the framework for institutional arrangements and oversight of AI systems, with entry into force planned for 2026.
The EU AI Act took effect in August 2024, with most rules applying in 2025–2026. This phased implementation offers Slovak companies limited but crucial time to prepare. Early investment in compliance systems, workforce training, and regulatory guidance will be essential to ensure full alignment with the Act. In addition, clear regulatory guidance is needed to clarify how the EU AI Act interacts with sector-specific rules in areas such as health, manufacturing, and mobility (OECD, 2026[9]).
Compliance with the EU AI act is likely to involve substantial costs and slow AI roll-out. Estimates from the EU’s impact assessment (European Commission: Directorate-General for Communications Networks, 2021[33]) indicate that compliance costs could be substantial, posing particular challenges for SMEs and startups that often lack financial resources and dedicated compliance staff. Without targeted support, small companies are likely to face a high compliance burden that could further slow innovation. From this perspective, the transposition of the EU AI Act into national legislation should prioritise simplification and avoid adding (through so-called “gold-plating”) to the EU-mandated requirements. The Digital Innovation Hubs should serve as resources to help reduce compliance costs for SMEs by assisting them in meeting legal and regulatory requirements.
At the same time, Slovakia’s ability to experiment and innovate in AI is constrained by the absence of operational regulatory “sandboxes”. Sandboxes are controlled environments where companies can test AI solutions under the supervision of authorities while being protected from legal risk (OECD, 2023[34]). AI regulatory sandboxes are explicitly promoted by the EU AI Act as a tool to improve legal certainty and foster innovation: they allow AI systems to be developed and tested with regulatory guidance before market release. The lack of sandboxes has deprived Slovak companies from a potentially crucial tool for experimentation. To address this gap, the Slovak Republic is currently preparing to launch a sandbox in August 2026, with a preparatory phase planned for end 2025, in line with the Digital Transformation Action Plan (2023–2026) and as envisaged in the forthcoming national AI legislation. The AI Sandbox is intended to align Slovakia’s AI ecosystem with the EU AI Act and create a safe testbed for AI developers. Its timely implementation will be critical to support innovation and helping smaller companies navigate the new regulatory framework. Widening the scope of the AI Sandbox to capture AI-based applications in other sectors may also contribute to a positive feedback loop between development of the technology and its applications (Calvino, Haerle and Liu, 2025[2]).
The rollout of AI in financial services requires attention to the specific regulatory framework applicable to the sector and risks that can emerge. As other forms of digital financial innovation, AI could amplify risks including in the areas of market integrity, consumer and investor protection (with “robo-advisors”) and financial stability (Nassr, 2025[35]). On the other hand, financial regulations can also result in sector-specific barriers to the deployment of AI technologies. This situation creates a case for financial policy makers to regularly review regulatory frameworks to remove unnecessary barriers that can shackle productivity gains while maintaining safeguards that guarantee financial stability and privacy (OECD, 2024[36]).
Another critical headwind for AI adoption and innovation in Slovakia are the significant gaps that persist in the availability of qualified professionals with expertise in AI. This includes both low digital and AI skills, and retention of skilled workers. Insufficient digital skills, including those related to AI, remain a key barrier to the uptake of digital technologies in Slovak businesses (Bendová, 2025[37]). Digital skills and those related to information communication technology (ICT), which provide the foundation for the development and use of AI (Calvino and Fontanelli, 2023[4]), have decreased over time. Managerial skills are also important for AI adoption, as AI must be integrated into existing legacy systems and workflows, requiring significant organisational change in addition to specialised expertise (Calvino et al., 2022[38]).
3.7.1. Companies find it difficult to hire workers with the skills needed for AI
Labour shortages in ICT roles remain high in Slovakia, where on average the number of vacancies per employed person in ICT jobs were 142% above the national average, exceeding the OECD average of 117% (OECD, 2024[39]). The overall share of ICT specialists in the workforce declined slightly, from 4.3% in 2023 to 4.2% in 2024 and was of 0.6 percentage points lower than the EU average of 4.8%. In 2024, only 51.3% of Slovaks had basic digital skills, compared to 55.6% across the EU (European Commisssion, 2024[40]).
In 2023, over 44% of Slovak businesses identified a lack of employee digital skills as a key barrier to adopting digital tools (Krištofičová, 2024[41]). In 2024, the main obstacles noted by companies to digital transformation included lack of digital skills, financial constraints, weak change management and the absence of digital leadership. Digital skill shortages were noted to be more pronounced in advanced areas such as AI (59%), followed by programming (36%) and data analysis (35%) (Bendová, 2025[37]). In 2022, over 60% of manufacturing companies reported insufficient digital skills among hired graduates. Similarly, 40% of businesses hiring for administrative positions indicated that recent graduates lacked the expected digital competencies, highlighting the need for better alignment between academic training and industry requirements (Kešelová et al., 2022[42]).
3.7.2. Higher education in relevant fields is low but expanding
Interest in ICT studies has increased over time. In 2023, about 3.9% of master graduates and 6.1% of bachelor graduates from Slovak universities completed degrees in the field of ICT, an increase of approximatively 60% and 90%, respectively, since 2015 (Figure 3.9 Panel A). In 2023, about 20% of bachelor’s and of master’s graduates had completed degrees in the field of STEM. These fields provide essential foundational skills for AI-related studies and careers, supporting the development of a digitally skilled workforce (OECD, 2025[43]). In addition, similar to OECD countries on average, Google trends data show an overall higher interest for AI education, possibly demonstrating growing awareness of the importance of AI-related skills (OECD.AI, 2025[44]). Nevertheless, by international comparison, relatively few young people in each age cohort are getting bachelors in STEM fields especially the ones most directly relevant to AI (Figure 3.9 Panel B).
Note: Category "other STEM" refers to the following educational fields: "natural sciences, mathematics and statistics" and "engineering, manufacturing and construction".
Sources: OECD (2025[43]); OECD Education Statistics database; OECD Population Statistics database: and OECD calculations.
Expanding higher-education programmes in computer science and other AI-relevant fields is a matter of priority. Since 2024, a national programme aims at strengthening links between secondary schools with specialisation in AI-relevant fields and corresponding higher-education institutions. From September 2026, a new international master-level program will be launched at the Slovak Technical University in Bratislava and the Technical University in Košice in collaboration with technological companies and foreign universities. For broad AI use across the economy, the acquisition of digital competencies should also be weaved into other study programmes.
3.7.3. Facilitating transitions to AI occupations
Most career transitions into AI occupations come from non-AI occupations. Over the last five years, on average 65.1% of LinkedIn members based in Slovakia who moved into AI jobs came from non-AI roles (OECD.AI, 2025[45]). This number, higher than that of neighbouring countries, suggests substantial upskilling and reskilling into the field of AI. It highlights that, while there are currently low levels of AI skills in the population, the local workforce is increasingly adapting and moving into AI-linked occupations.
Informal learning appears to play a large role in AI upskilling. Close to 89% of the respondents to the Stack Overflow survey of developers had noted they had learned how to code online, compared to 50% through the formal education system, and 34% during job training. This is compared to the overall average for survey respondents of 76% learning through online resources, 45% through education and 42% via job training. Slovak respondents therefore relied considerably more on self-directed online resources than the education system (Figure 3.10). In-firm-provided training can benefit both the enterprise and its employees, leading to higher productivity and firm performance, as well as increased wages and career advancements opportunities for workers (OECD, 2021[46]).
Incentives and policy interventions can help address the under-provision of in job training. Governments should encourage employers to provide more training mechanisms to support adoption of AI in the workplace (OECD, 2023[47]). Over time, these efforts should benefit from the entry in the labour force of young workers having learned basic computer-science and AI concepts as part of their secondary education following the 2023 curriculum reform.
Despite relatively low adoption at the company level, the labour force in Slovakia still reports interest and exposure to AI. About half of the Slovak workforce is highly exposed to AI technologies across different sectors, including education, professional services, finance, information and communication, and the public administration (IMF, 2025[48]). The share of workers exposed to AI deployment, however, varies across regions, education levels and gender, with female workers especially affected. On average, 24.5% of the workforce is exposed to tasks that could be affected by generative AI, with regional exposure ranging from 16.5% to 35.8%. An estimated 35.8% of workers face a high risk of task automation, compared with 27% across OECD countries (OECD, 2024[39]; OECD, 2023[47]).
Distribution of coding learning resources among survey respondents based in Slovakia
Notes: Aggregate demographic information from survey respondents is leveraged to build indicators and identify trends related to profession, education, salary and age of AI developers. The 2024 survey included 248 responses from Slovakia-based participants, out of a total of 65,437 respondents across 185 countries (OECD.AI, 2025[49]).
Source: OECD.AI (2025[49]), using data from Stack Overflow available at: https://oecd.ai/data?selectedArea=ai-demographics.
This exposure to AI, combined with automation risks, highlights the urgency of equipping the workforce with stronger digital skills and of managing labour market transitions. The 2030 Digital Transformation Strategy for Slovakia recognises a shortage of labour with advanced digital skills or with experience in the use of technologies and, a possible insufficient technical education in Slovakia (MIRRI, 2019[50]). Without targeted measures, the gap between jobs likely to be disrupted by AI and the skills needed to adapt could widen, resulting in structural mismatches in the labour market. To address this challenge, the strategy includes measures to retain and attract talent, both from Slovakia and from abroad.
Slovakia has launched several initiatives to strengthen the adoption of AI-related skills including the Plan for Responsible Use of AI in Education (Ministry of Education, Research, Development and Youth, 2025[16]). These efforts are integrating foundational competencies enabling fluent AI use, including critical thinking and problem solving, alongside digital skills in primary and secondary education (Ministry of Education, Research, Development and Youth, 2025[17]; Borgonovi et al., 2023[51]). A central initiative in this respect is the “Support for the Digital Transformation of Education” (DiTEdu) project that aims to incorporate topics such as introduction to neural networks, artificial intelligence, algorithmic thinking and programming languages for AI into study materials across primary and secondary schools (OECD, 2026[9]).
An important building block was the reintroduction of mathematics as part of the core curriculum until the end of secondary education under the 2023-2027 educational reform (Ministry of Education, Research, Development and Youth, 2025[52]). The reform broadly strengthened STEM foundations across secondary and vocational education. The availability of teachers with the required competencies, access to high-speed connectivity and purchase of digital devices remain hurdles for many schools. In addition, the Ministry of Investments, Regional Development and Informatisation of the Slovak Republic (MIRRI) has launched AI literacy and ethics training programmes for civil servants (European Commission, 2021[53]).
Policymakers should continue to promote actions to strengthen the development of AI skills. This includes encouraging firms to provide in-job training and integrating AI skills into education systems, such as vocational education and lifelong-learning programmes. Such measures can provide opportunities to skill, reskill and upskill workers in AI-relevant fields throughout their education and careers. Promoting micro-credentials as part of publicly funded training programmes can be an effective part of the toolbox to facilitate rapid skill development in the tech sector (OECD, 2024[54]).
3.7.4. Retaining AI-fluent workers is a challenge
In parallel, the Slovak Republic faces challenges in retaining and attracting AI-skilled labour. The emigration and brain drain that have been major drivers of skill shortages and skill mismatches across the labour market [see Chapter 4 and OECD (2020[55])] have also impacted the digital sector (ITA, 2024[56]).
Lower salaries for skilled AI workers are likely contributing to Slovakia’s brain drain. Among respondents of the Stack Overflow survey who reported coding in their professional occupation, 61% had an advanced degree (e.g. masters, PhD degrees) and nearly 11% had a bachelor’s degree (OECD.AI, 2025[49]). The average salary of the respondents per year was of USD 47,000 close to 29% higher than the average annual wage in Slovakia in 2024 of USD 36,105 in PPP (OECD, 2021[57]; 2025[58]). However, when compared internationally, Slovak AI professionals earn considerably less than their peers in neighbouring countries. For example, the average Slovak salary of respondents is about 24% lower than reported by developers and data scientist based in Poland and 15% lower than those in Czechia (OECD.AI, 2025[49]).
Furthermore, the tax wedge in the Slovak Republic remains high compared with the OECD average and neighbours especially for highly skilled workers (see Chapter 4). Combined with lower average salaries, these wage gaps suggest that even well-educated Slovak AI specialists may seek more lucrative opportunities abroad, reinforcing the country’s talent outflow.
In the face of these challenges, the authorities are implementing measures aiming to retain domestic and attract foreign AI talent. These include scholarships as well as English-language higher-education programmes, such as the above-mentioned international Master of Science in AI. These measures also cover other fields with competency shortages such as nuclear science, physics, electrical engineering, and mechanical engineering. To be most effective, these focussed initiatives need to be implemented together with broader efforts to lower the tax wedge faced by high-skilled workers and boost economic dynamism (see Chapter 2).
References
[37] Bendová, A. (2025), Prieskum: Digitalizuje tretina podnikov, slovenské malé a stredné firmy zaostávajú, https://industry4um.sk/tretina-podnikov-digitalizuje-slovenske-male-a-stredne-firmy-zaostavaju/ (accessed on 6 August 2025).
[51] Borgonovi, F. et al. (2023), “Emerging trends in AI skill demand across 14 OECD countries”, OECD Artificial Intelligence Papers, No. 2, OECD Publishing, Paris, https://doi.org/10.1787/7c691b9a-en.
[38] Calvino, F. et al. (2022), “Closing the Italian digital gap: The role of skills, intangibles and policies”, OECD Science, Technology and Industry Policy Papers, No. 126, OECD Publishing, Paris, https://doi.org/10.1787/e33c281e-en.
[4] Calvino, F. and L. Fontanelli (2023), “A portrait of AI adopters across countries: Firm characteristics, assets’ complementarities and productivity”, OECD Science, Technology and Industry Working Papers, No. 2023/02, OECD Publishing, Paris, https://doi.org/10.1787/0fb79bb9-en.
[2] Calvino, F., D. Haerle and S. Liu (2025), “Is generative AI a General Purpose Technology?: Implications for productivity and policy”, OECD Artificial Intelligence Papers, No. 40, OECD Publishing, Paris, https://doi.org/10.1787/704e2d12-en.
[27] Cremona, E. and P. Czyzak (2025), Grids for data centres: ambitious grid planning can win Europe’s AI race.
[30] Data Center Map (2025), Data centers, https://www.datacentermap.com/datacenters/.
[28] Epoch AI (2025), Data on AI models, https://epoch.ai/data/ai-models.
[20] European Comission (2025), Public Sector Tech Watch, https://interoperable-europe.ec.europa.eu/collection/public-sector-tech-watch.
[14] European Commission (2025), European Digital Innovation Hubs, https://digital-strategy.ec.europa.eu/en/policies/edihs.
[53] European Commission (2021), Slovakia AI Strategy Report, https://ai-watch.ec.europa.eu/countries/slovakia/slovakia-ai-strategy-report_en?utm_source=chatgpt.com#human-capital (accessed on 29 August 2025).
[33] European Commission: Directorate-General for Communications Networks, C. (2021), Study to support an impact assessment of regulatory requirements for Artificial Intelligence in Europe – Final report, Publications Office, https://op.europa.eu/en/publication-detail/-/publication/55538b70-a638-11eb-9585-01aa75ed71a1.
[40] European Commisssion (2024), Slovakia 2024 Digital Decade Country Report, https://digital-strategy.ec.europa.eu/en/factpages/slovakia-2024-digital-decade-country-report.
[5] Eurostat (2025), Statistics Explained: Use of Artificial Intelligence in Enterprises, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Use_of_artificial_intelligence_in_enterprises.
[1] Filippucci, F. et al. (2026), “AI meets trade: Global linkages and the cross-country distribution of the gains from AI”, OECD Artificial Intelligence Papers, No. 57, OECD Publishing, Paris, https://doi.org/10.1787/13081644-en.
[48] IMF (2025), “Slovak Republic”, IMF Staff Country Reports, Vol. 2025/073, p. 1, https://doi.org/10.5089/9798229005852.002.
[10] International Federation of Robotics (2024), World Robotics - Industrial Robots, https://ifr.org/wr-industrial-robots/.
[56] ITA (2024), Slovakia Country Commercial Guide, https://www.trade.gov/country-commercial-guides/slovakia-digital-economy (accessed on 4 August 2025).
[11] Kergroach, S. and J. Héritier (2025), “Emerging divides in the transition to artificial intelligence”, OECD Regional Development Papers, No. 147, OECD Publishing, Paris, https://doi.org/10.1787/7376c776-en.
[42] Kešelová, D. et al. (2022), The impact of robotization, automation and digitization on the labor market in the Slovak Republic. Results of the empirical survey, https://ivpr.gov.sk/wp-content/uploads/2024/01/Vyskumna_sprava_digitalizacia.pdf (accessed on 28 August 2025).
[41] Krištofičová, L. (2024), Digital and green skills will help halt Slovakia’s decline in competitiveness, https://digital-skills-jobs.europa.eu/en/latest/news/digital-and-green-skills-will-help-halt-slovakias-decline-competitiveness (accessed on 5 August 2025).
[8] Milanez, A. (2023), “The impact of AI on the workplace: Evidence from OECD case studies of AI implementation”, OECD Social, Employment and Migration Working Papers, No. 289, OECD Publishing, Paris, https://doi.org/10.1787/2247ce58-en.
[52] Ministry of Education, Research, Development and Youth (2025), Educational Reform in Slovakia 2023-2025.
[17] Ministry of Education, Research, Development and Youth (2025), Integration of AI into Slovak Education.
[16] Ministry of Education, Research, Development and Youth (2025), Plan for Responsible Use of AI in Education 2025-2027.
[15] MIRRI (2021), Národná koncepcia informatizácie verejnej správy Slovenskej republiky (The National Concept of Informatization of the Public Administration for years 2021-2026), https://mirri.gov.sk/wp-content/uploads/2021/12/Narodna-koncepcia-informatizacie-verejnej-spravy-2021.pdf.
[50] MIRRI (2019), 2030 Digital Transformation Strategy for Slovakia, https://mirri.gov.sk/wp-content/uploads/2019/10/SDT-English-Version-FINAL.pdf.
[3] OECD (2026), OECD.AI Index, OECD Publishing, Paris, https://doi.org/10.1787/32c01014-en.
[9] OECD (2026), Progress in Implementing the European Union Coordinated Plan on Artificial Intelligence (Volume 2): Uptake in High-Impact Sectors, OECD Publishing, Paris, https://doi.org/10.1787/3ac96d41-en.
[23] OECD (2025), Advancing the measurement of investments in Artificial Intelligence: objectives, definitions, measurement framework and estimates for the European Union, https://one.oecd.org/official-document/DSTI/DPC/AIGO(2025)2/en.
[13] OECD (2025), AI adoption by small and medium-sized enterprises: OECD discussion paper for the G7, OECD Publishing, Paris, https://doi.org/10.1787/426399c1-en.
[35] OECD (ed.) (2025), AI in finance: Balancing innovation and stability, https://oecd.ai/en/wonk/ai-in-finance-balancing-innovation-and-stability.
[58] OECD (2025), Average annual wages, https://data-explorer.oecd.org/s/2oy (accessed on 28 August 2025).
[43] OECD (2025), Distribution of enrolled students, new entrants and graduates by field of education, https://data-explorer.oecd.org/s/2p0 (accessed on 1 August 2025).
[12] OECD (2025), Generative AI and the SME Workforce: New Survey Evidence, OECD Publishing, Paris, https://doi.org/10.1787/2d08b99d-en.
[18] OECD (2025), Governing with Artificial Intelligence: The State of Play and Way Forward in Core Government Functions, OECD Publishing, https://doi.org/10.1787/795de142-en.
[19] OECD (2025), Progress in Implementing the European Union Coordinated Plan on Artificial Intelligence (Volume 1): Slovakia Country Note, OECD Publishing, https://www.oecd.org/en/publications/progress-in-implementing-the-european-union-coordinated-plan-on-artificial-intelligence-volume-1_6d530a88-en/slovak-republic_5b824a40-en.html.
[26] OECD (2025), “Sharing trustworthy AI models with privacy-enhancing technologies”, OECD Artificial Intelligence Papers, No. 38, OECD Publishing, Paris, https://doi.org/10.1787/a266160b-en.
[54] OECD (2024), Bridging Talent Shortages in Tech: Skills-first Hiring, Micro-credentials and Inclusive Outreach, Getting Skills Right, OECD Publishing, Paris, https://doi.org/10.1787/f35da44f-en.
[39] OECD (2024), Job Creation and Local Economic Development 2024 - Country Notes: Slovak Republic, https://www.oecd.org/en/publications/job-creation-and-local-economic-development-2024-country-notes_ad2806c1-en/slovak-republic_e69fac3d-en.html.
[36] OECD (2024), “Regulatory approaches to Artificial Intelligence in finance”, OECD Artificial Intelligence Papers, No. 24, OECD Publishing, Paris, https://doi.org/10.1787/f1498c02-en.
[31] OECD (2023), “A blueprint for building national compute capacity for artificial intelligence”, OECD Digital Economy Papers, No. 350, OECD Publishing, Paris, https://doi.org/10.1787/876367e3-en.
[47] OECD (2023), OECD Employment Outlook 2023: Artificial Intelligence and the Labour Market, OECD Publishing, Paris, https://doi.org/10.1787/08785bba-en.
[34] OECD (2023), Regulatory sandboxes in artificial intelligence, OECD Digital Economy Papers, No. 356, OECD Publishing, Paris, https://doi.org/10.1787/8f80a0e6-en.
[7] OECD (2022), “OECD Framework for the Classification of AI systems”, OECD Digital Economy Papers, No. 323, OECD Publishing, Paris, https://doi.org/10.1787/cb6d9eca-en.
[57] OECD (2021), OECD Economic Outlook, Volume 2021 Issue 1, OECD Publishing, Paris, https://doi.org/10.1787/edfbca02-en.
[46] OECD (2021), Training in Enterprises: New Evidence from 100 Case Studies, Getting Skills Right, OECD Publishing, Paris, https://doi.org/10.1787/7d63d210-en.
[55] OECD (2020), OECD Skills Strategy Slovak Republic: Assessment and Recommendations, OECD Skills Studies, OECD Publishing, Paris, https://doi.org/10.1787/bb688e68-en.
[49] OECD.AI (2025), Live data, AI demographics, https://oecd.ai/en/data?selectedArea=ai-demographics (accessed on 4 August 2025).
[45] OECD.AI (2025), Live data, AI jobs and skills, https://oecd.ai/en/data?selectedArea=ai-jobs-and-skills&selectedVisualization=career-transitions-into-ai-occupations-all-countries-2023 (accessed on 5 August 2025).
[44] OECD.AI (2025), Live data, AI search trends, https://oecd.ai/en/data?selectedArea=ai-search-trends (accessed on 4 August 2025).
[29] OECD.AI (2025), Live data, AIKoD database, https://oecd.ai/en/data?selectedArea=ai-models-and-datasets.
[22] OECD.AI (2025), Live data, Preqin, https://oecd.ai/en/data?selectedArea=investments-in-ai-and-data.
[21] OECD.AI (2025), Live data, Scopus AI research, https://oecd.ai/en/data?selectedArea=ai-research&selectedVisualization=16738.
[25] Office of the Deputy Prime Minister for Investments and Informatisation of Slovakia (2022), Strategy for the Digital Transformation of Slovakia 2030.
[6] SARIO (2025), Shared Service & Business Process Outsourcing Centers in SLOVAKIA, Slovak Investment and Trade Development Agency, https://www.sario.sk/sites/default/files/sario-ssc-bpo-centers-in-slovakia-2025-02-28.pdf (accessed on 14 October 2025).
[32] TOP500 (2025), , https://top500.org/lists/top500/2025/06/.
[24] VAIA, P. (2025), Inovačný voucher ako kľúč k modernizácii, https://vaia.gov.sk/wp-content/uploads/2025/06/Analyza-voucherovej-podpory.pdf.