The European Union Coordinated Plan on Artificial Intelligence is a strategic initiative developed by the European Commission and EU Member States to promote development, deployment and use of AI technologies across the European Union. Within this framework, this chapter looks at how EU Member States are building strategic leadership in five high-impact sectors. In the first section, it examines policies to support AI applications for climate and the environment. The next two sections discuss national initiatives leveraging AI to improve healthcare and drive transformation in the public sector. The fourth section analyses efforts to make mobility smarter, safer, and more sustainable through AI. Finally, the chapter reviews measures promoting the adoption of AI-enabled tools and solutions in agriculture.
Progress in Implementing the European Union Coordinated Plan on Artificial Intelligence (Volume 1)
Abstract
The EU Coordinated Plan on AI is a strategic initiative to promote development, deployment and use of AI technologies across the European Union. It represents a joint commitment between the European Commission and EU Member States to maximise the impact of investments in AI, foster synergies and encourage co-operation across the European Union. The plan outlines a series of concrete actions to facilitate investment decisions, aligning AI policy in the European Union to remove fragmentation. It also aims to contribute to strengthening the global position of the European Union regarding the development and adoption of human centric, sustainable, secure, inclusive and trustworthy AI technologies and applications.
Within this framework, this chapter looks at how EU Member States are building strategic leadership in five high-impact sectors. In the first section, it examines policies to support AI applications for climate and the environment. The next two sections discuss national initiatives leveraging AI to improve healthcare and drive transformation in the public sector. The fourth section analyses efforts to make mobility smarter, safer, and more sustainable through AI. Finally, the chapter reviews measures promoting the adoption of AI-enabled tools and solutions in agriculture.
The EU Coordinated Plan on AI identifies AI as a key enabler for achieving climate neutrality and advancing environmental sustainability. The 2021 update to the plan emphasises the need to harness AI for emissions reduction, circular economy innovation and natural resource protection. At the same time, it recognises the need to address the environmental footprint of AI systems themselves: the training of large language models can emit hundreds of tonnes of carbon dioxide, and their daily operation adds to the energy demands of global data centres (George, George and Martin, 2024[1]; Ruf, 2024[2]). These impacts extend across the full life cycle of AI compute, including hardware manufacturing, data storage and end-of-life disposal (Ligozat et al., 2021[3]; OECD, 2022[4]).The goal is to ensure that AI contributes to the objectives of the European Green Deal, both as a technological enabler of climate action and as a digital technology that must itself be sustainable.
To this end, EU Member States are encouraged to take the following actions:
Share results from national efforts on ‘green AI’ and climate actions, share best practices with other Member States and, on the basis of their experiences suggest cross-border projects, outreach efforts and action that could be taken at European level.
Share locally available expertise and know-how through the EDIH network which can support training and knowledge-sharing activities.
Support the inclusion of a ‘green AI’ component in university and higher education AI curricula and other AI training courses and programmes; and
Work with national ICT and other sectoral stakeholders, including standardisation bodies towards defining deployment guidelines and standardised assessment methodologies to support ‘green AI’ in areas such as smart grids, precision farming, and smart and sustainable cities as well as communities.
These coordinated efforts aim to ensure that AI becomes not only a catalyst for Europe’s green transition, but also a technology developed and deployed in line with the Union’s broader sustainability objectives.
In the climate and energy domains, AI applications are increasingly used to enhance modelling, support mitigation and adaptation planning, and improve the efficiency of energy systems. AI can help anticipate extreme weather events, optimise energy generation and storage, and use energy efficiently across sectors through smart resource management (Leal Filho et al., 2022[5]). In energy systems, AI-driven solutions improve load forecasting, balance supply and demand, support integration of renewable energy and reduce losses in transmission and distribution (Chen et al., 2023[6]). AI also underpins digital twin simulations and smart grids that can enhance real-time decision making, automate grid management and promote decentralised energy production (Kar, Choudhary and Singh, 2022[7]).
AI can process vast volumes of environmental and geospatial data, offering unique opportunities for climate modelling, emissions monitoring and biodiversity tracking. For instance, AI techniques are now used to monitor land-use change, track deforestation and model hydrological systems for early flood warnings and drought prediction (Chen et al., 2023[6]; Leal Filho et al., 2022[5]). Moreover, public and private actors are increasingly employing AI tools to support carbon accounting, assess corporate climate risk, and improve the design and enforcement of environmental regulation (Muench et al., 2022[8]; Barnabas and Owen, 2025[9]).
More than two-thirds of EU Member States have adopted policies to support AI applications for climate and environmental sustainability
Most EU Member States have adopted policies to support AI applications for climate and environmental sustainability. Most of these applications foster development of AI-based solutions for sustainability. Fifteen countries reported initiatives to apply AI to areas such as energy efficiency, emission reduction, waste management and disaster resilience. Only seven EU Member States have launched dedicated initiatives to gather environmental data and enhance monitoring capabilities, and only six EU Member States have put forward policies to reduce the environmental impact of AI (Figure 5.1).
Source: Data reported by EU Member States through the survey and interviews.
More than half of EU Member States are investing in AI to tackle environmental challenges
Fifteen EU Member States have introduced policy initiatives leveraging AI to advance sustainability objectives (Figure 5.2). These initiatives span diverse domains, including energy optimisation, waste reduction, disaster forecasting and urban climate resilience, highlighting the potential of AI to address environmental challenges. However, the distribution of efforts remains uneven across the European Union.
Source: Data reported by EU Member States through the survey and interviews.
Several countries have launched comprehensive programmes that integrate AI across multiple environmental application areas.
Germany is among the frontrunners with several large-scale initiatives. With a budget of EUR 150.9 million, AI-Lighthouses for Environment, Climate, Nature and Resources funds AI solutions for energy and resource efficiency, biodiversity protection and sustainable mobility (BMUV, 2024[10]). The Application Lab for AI and Big Data, operated by the German Environment Agency, leverages AI for environmental research, enforcement and administrative innovation (UBA, 2024[11]). Meanwhile, the AI Idea Workshop for Environmental Protection supports civil society organisations in co-developing AI applications for environmental challenges (KI-Ideenwerkstatt, 2025[12]).
Greece is applying AI-based tools in six urban centres under the EU 100 Climate Neutral Cities mission, targeting emissions reduction, waste management and energy optimisation (BBK, 2024[13]). Italy has scaled up simulation capacity through the Cassandra supercomputer, enabling new AI-powered climate science at the Euro-Mediterranean Center on Climate Change. The Netherlands, through its EUR 189 million AiNed programme, launched AI for Energy and Sustainability, creating AI Innovation Labs to support the energy transition (AiNed, 2024[14]). Portugal’s National Strategy for Smart Territories brings together municipalities, academia and industry to use AI, Internet of Things (IoT) and the fifth generation of cellular network technology (5G) for smarter water, mobility and energy systems (AMA, 2023[15]).
Spain funds two National Artificial Intelligence Strategy (ENIA) Chairs, one dedicated to sustainability and a decarbonised society, the other to green algorithms. Both focus on development and deployment of AI for environmental applications. The SMASH programme in Slovenia explores the use of machine learning in climate modelling, extreme weather analysis and air pollution source identification (SMASH, 2024[16]). The country is also participating in the EU NetZeroCities pilot programme, with the cities of Ljubljana, Kranj and Velenje, deploying AI and digital platforms to support urban decarbonisation strategies through improved data integration and planning (NetZeroCities, 2024[17]).
Energy efficiency
Multiple EU Member States are deploying AI-driven solutions to optimise energy consumption, grid resilience and renewable integration.
Belgium is advancing AI‑powered energy management through the Coordination Mechanisms for the Sharing of Energy through Proxies (COOMEP) initiative, which enables households to optimise energy use via AI-based decision-making systems (VUB, 2024[18]). The country is also implementing the EUR 247 000 SIRIUS Simulator Project, an AI tool supporting electric grid planning (Europol, 2024[19]). This is a predictive maintenance initiative for wind turbine parks under the Prognostic Health Management and Improving Energy Production of Wind Turbine Parks project (Flanders AI Research Program, 2024[20]). Another Belgian initiative, Digital Twin in Support of Sustainable and Resilient Energy Systems, applies AI-powered simulations to improve grid stability and sustainability (VUB, 2024[18]).
Ireland is integrating AI into urban energy planning through Next Generation Energy Systems (NexSys). This EUR 18.3 million research initiative explores AI-based optimisation of neighbourhood energy distribution (NexSys, 2022[21]). Italy’s Transition Plan 5.0 incorporates AI into its EUR 6.3 billion national sustainability strategy. In so doing, it supports energy-efficient digital infrastructure and smart resource management. Latvia is leveraging AI to enhance energy sector efficiency through the EUR 5.59 million I‑NERGY project (REA, 2021[22]) and a EUR 1.57 million research collaboration. The latter falls under the Swiss Latvian Cooperation Programme – Pre-defined Programme Component in Information and Communication Technology (ICT) and Smart Energy – which focuses on AI applications in smart grids and energy storage. In Poland, New Energy (Nowa Energia) programme, backed by EUR 590 million, supports AI-driven innovation in smart city energy systems and decentralised energy clusters (Government of Poland, 2025[23]).
Weather forecasting and risk reduction
EU Member States are increasingly applying AI in the context of weather disaster forecasting and risk reduction. France and Germany co-fund the CONTRAILS project to assess and reduce the climate impact of aircraft condensation trails through hybrid AI and physical modelling (Deutscher Wetterdienst, 2025[24]). In addition, the RenovAIte initiative applies AI to optimise infrastructure renovation (RenovAIte, 2025[25]). Bulgaria is part of the EU-funded ForeSight project, which uses real-time data and AI to predict and manage wildfires and urban emergencies. Latvia is developing an AI-powered system to forecast storm‑induced damage to power grids, improving resilience against extreme weather events (Sadales tīkls, 2025[26]). Slovenia has deployed HIDRA, which uses deep neural networks to predict sea levels and coastal flooding risks.
Business applications
Several EU Member States are supporting AI adoption in businesses to foster sustainability. Germany launched the EUR 15.2 million Green-AI Hub Mittelstand, which supports small and medium-sized enterprises (SMEs) in implementing AI to reduce energy and resource consumption (BMUV, 2024[27]). Smart Attica in Greece and sustAIn.brussels in Belgium act as European Digital Innovation Hubs (EDIHs). They offer training and technical assistance to enterprises building environmentally focused AI solutions (sustAIn.brussels, 2024[28]; Smart Attica, 2024[29]). For its part, Slovenia has included AI-based pilot projects in its strategy to increase renewable energy uptake in the economy. This includes peer-to-peer energy marketplaces and improved emissions monitoring through digital tools (Slovenian Ministry of Economic Development and Technology, 2022[30]).
Waste management
AI is also being applied to improve waste management and recycling efficiency. Germany has established the Application Hub for a Circular Economy for Plastic Packaging through AI Methods. This EUR 30 million initiative supports research into AI-based solutions to reduce plastic waste and enhance circularity (AI Hub Plastic Packaging, 2024[31]). Latvia is advancing automated waste sorting through the WinGo Deposit system. It uses machine vision and neural networks to classify recyclable materials such as polyethylene terephthalate (PET) bottles, metal cans and batteries, with plans to expand to additional waste types (Labs of Latvia, 2023[32]).
Environmental subfields
Some countries reported AI initiatives in specific environmental subfields. Bulgaria is building a national real-time water management system that integrates AI to monitor precipitation, snowmelt and river flows for better flood risk assessment. These cases illustrate the diversity of potentially using AI for environmental challenges, even if such uses remain the exception rather than the norm among EU Member States. Croatia has adapted the ATMOSYS Air Quality Management System, originally developed in Belgium, to forecast localised air quality using machine-learning tools. The system supports compliance with EU air legislation at both the regional and urban levels. France supports marine science through the OcéanIA Challenge, a research initiative developing AI and mathematical modelling tools to study the ocean’s role in climate regulation and biosphere preservation (Inria, 2025[33]). Malta is applying AI to cultural heritage preservation through the MEGALITH project, which uses climate modelling and simulation data to predict stone degradation in megalithic temple sites (University of Malta, 2023[34]).
Environmental monitoring
Seven EU Member States have launched dedicated initiatives for environmental monitoring and data-driven policymaking. Data collection and monitoring are essential for informed policymaking and efficient resource management. AI is increasingly being leveraged to improve the accuracy, efficiency and integration of environmental data, enabling real-time monitoring, predictive analytics and cross-sector data sharing. Seven of 27 EU Member States have launched AI-driven initiatives focused on strengthening environmental data systems (Figure 5.1). These projects range from energy and emissions monitoring to geospatial mapping and data infrastructure development.
Countries have started to introduce centralised platforms to integrate and enhance access to environmental data. Austria has developed the Climate Change Cockpit, a national platform designed to consolidate fragmented environmental data, particularly for climate-sensitive sectors like tourism (InnoDays, 2024[35]). Similarly, in Slovenia, the Data Warehouse for the Creation of the Energy Information System EnergIS is establishing a scalable data infrastructure to integrate energy-related datasets and improve decision making for climate and energy policies (Slovenian Ministry of Digital Transformation, 2023[36]).
AI is also supporting real-time environmental monitoring and geospatial analysis, providing critical insights into emissions, energy usage and environmental conditions. The REASSURE project in Belgium provides an AI-driven data analytics toolkit to optimise the reliability and sustainability of energy systems through predictive modelling (VUB, 2024[37]). In Bulgaria, the Geospatial Data for Environment integrates AI‑powered geographic information services for tracking biodiversity, water resources, land cover and pesticide use. In Finland, the Visiiri Project, part of the Green ICT ecosystem, assesses the environmental impact of the country’s ICT sector while promoting AI-enabled solutions for sustainability. Lithuania is using AI in the Improvement of Greenhouse Gas Accounting programme to assess land-use changes and refine emissions monitoring methodologies, ensuring compliance with evolving EU climate regulations. Malta is piloting AI for Better Utilities, which applies AI to analyse water and energy consumption patterns, with the aim of improving the resilience and efficiency of its utility networks. DS2 – DataSpace, DataShare 2.0 project – is advancing cross-sector data sharing in Slovenia by linking multiple environmental data sources, facilitating real-time monitoring of air pollutants and greenhouse gas emissions (DS2, 2023[38]).
These initiatives mark an important step towards strengthening the environmental data ecosystem in the European Union. Expanding use of AI in emissions monitoring, geospatial analytics and integrated data platforms will be crucial for enhancing climate adaptation strategies and driving data-informed environmental policies across the European Union.
Six EU Member States are pursuing efforts to reduce the environmental footprint of AI
Six EU Member States have introduced structured initiatives to mitigate the energy and resource consumption of AI (Figure 5.1), highlighting a significant gap in policy action. More than three-fifths of Member States have not yet adopted measures to address the environmental footprint of AI, leaving efforts concentrated in a few countries. Reported initiatives focus on improving the energy efficiency of AI models and data centres, integrating AI into broader sustainable digitalisation strategies and promoting research on frugal AI. While these initiatives represent important progress, further action is needed to ensure AI development aligns with EU climate and energy goals, particularly as demand for high-performance computing continues to rise.
Some governments have launched strategy documents to foster sustainable AI development. Finland has outlined measures to reduce the environmental footprint of the ICT sector, including AI technologies, in its Climate and Environmental Strategy for the ICT Sector (Finnish Ministry of Transport and Communications, 2021[39]). Meanwhile, the Netherlands has developed the Sustainable Digitalisation Action Plan. It sets out 44 targeted actions to align digitalisation efforts with sustainability goals, including reducing AI-related energy and water consumption. Similarly, Spain has introduced the National Green Algorithms Program. It promotes “green-by-design” AI models that integrate environmental sustainability variables and encourage synergies between the green and digital transitions (PNAV, 2025[40]).
Investments in frugal, energy-efficient AI and measures to minimise the environmental footprint of AI infrastructure are gaining momentum. The Green Data Processing and Storage initiative in Denmark is developing best practices for energy-efficient data centres and AI applications, with a focus on procurement guidelines and energy efficiency assessments (Digitaliseringsstyrelsen, 2025[41]). France is supporting multiple projects in this area. The General Framework for Frugal AI provides a methodology for assessing and mitigating AI’s environmental impact (Ecolab, 2024[42]). In addition, PEPR IA Projects on Frugal and Embedded AI advance research in energy-efficient AI models. France has also funded research into reducing the energy consumption of cloud computing, with projects such as Pushing Low-carbon Services towards the Edge and End-to-end Eco-design of a Cloud to Reduce its Environmental Impact, in collaboration with OVHcloud and Qarnot Computing. Slovenia has committed to using renewable energy sources to power new supercomputing infrastructure in Arnes, with plans to integrate hydroelectric power and repurpose waste heat from the system to heat parts of the town of Maribor. In Spain, Artificial Intelligence for a Sustainable Energy Transition also contributes to this effort by exploring AI-enabled efficiency improvements across the energy value chain.
With the increasing availability of high-quality health data, AI holds immense potential to transform healthcare by optimising health service management, improving clinical decision making, enhancing personalised medicine and facilitating predictive medicine. Beyond improving the quality of care, these advancements can collectively contribute to advance the quintuple aim for healthcare by reducing cost of care, improving individual and population health, enhancing patients’ experience, increasing providers’ satisfaction and promoting health equity (OECD, 2019[43]). The ability of the European Union to fully harness the benefits of safe, secure and trustworthy AI in healthcare depends upon the capacity of EU Member States to strategically align the approach for the responsible adoption of AI and related investments, as well as to implement and harmonise data governance frameworks.
Against this background, the EU Coordinated Plan on AI encourages Member States to:
Take actions to increase the quality and semantic interoperability of health data, which is fundamental for the development and use of AI.
Develop actions and support initiatives to increase medical professionals’ understanding and acceptance of digital technology to accelerate adoption of AI-based systems in the medical field.
Implement recommendations that promote the eHealth upskilling of healthcare workers and agree on common European quality indicators for continued medical education.
Advance the ‘1+ million genomes’ initiative possibly through their national recovery and resilience plans, including as a multi-country project.
Support investments in secondary use of health data, including for AI, using, for example, RRF funding.
Take action to facilitate the integration of innovative AI-based systems (e.g. machine learning, autonomous systems, conversational agents, big data, robotics) in health and care facilities such as hospitals and care homes, and notably when the digitalisation of the health systems has been outlined in the national recovery and resilience plans.
Support EDIHs specialised in medical technologies and eHealth in order to help regional/national health systems and industry in their research efforts to provide better treatments and advances towards beating the coronavirus; and
Work with national, regional and international standardisation bodies to formulate towards defining and setting common standards, including on issues such as security, safety, privacy, interoperability, in an effort to update existing standards for AI for health.
EU Member States still struggle to leverage AI benefits. European financial incentives such as the Recovery and Resilience Facility (RRF) and targeted funding for the 1+ Million Genomes (1+MG) aim to develop a unified AI in healthcare approach. However, EU Member States continue to face challenges in fully leveraging AI benefits due to several factors. These comprise fragmented policies across borders; varying interpretation and application of EU legislation; and limited co‑ordination between country-level initiatives. This divergence risks limiting opportunities for cross-border knowledge sharing and slowing progress in innovation. By fostering a robust collaboration – such as through the European Health Data Space (EHDS) – the European Union could create a more resilient, safe and innovative AI health sector. Such a sector could build on the individual strengths of EU Member States, while reinforcing the competitive edge of the European Union in the global health landscape.
EU Member States are making significant strides in AI-driven healthcare innovation
Although EU Member States are generally furthering AI-driven healthcare innovation, the complexity of health data and varied regulatory environments demand greater collective action and co-operation. Establishing a robust foundation for AI, built upon compatible data-driven policies and harmonised technical and semantic standards, is crucial to enabling AI to thrive across borders. These pillars are necessary not only for regulatory consistency but also to facilitate cross-border collaboration to enable the efficient sharing of leading practices. Building shared practices such as interoperable data systems, regulatory frameworks and ethical standards will enable countries to innovate while maintaining high security and privacy safeguards.
As country-level and EU-wide initiatives are developed and lessons are drawn from implementation, they can inform and update policies and regulations, leading to a stronger and more harmonised foundation. This process, in turn, incentivises EU Member States to develop robust enablers, such as comprehensive and quality health datasets, to implement more effective and relevant initiatives. In this way, they can achieve a greater impact on the quintuple aim for healthcare. Skipping these foundational steps risks further cross-national fragmentation, embeds long-standing data integration challenges and ultimately reduces the potential impact of AI on improving health outcomes.
The responses of EU Member States on the thematic areas highlight varying levels of maturity in establishing robust health data governance frameworks and technical infrastructure, particularly in light of the 2021 updates to the EU Coordinated Plan on AI. In one trend, EU Member States such as in Finland and the Netherlands – with reliable digital infrastructure, interoperable health systems and co‑ordinated policies across governance, technology and people capacity – are further along in AI development (Box 5.1). However, this advancement depends highly on the structure and co‑ordinated mechanisms within each health system. EU Member States have begun to scale AI across existing frameworks in specific use-cases. In addition, there has been limited assessment of AI acceptance among the public and health professionals. However, significant efforts are devoted to expanding medical curricula to ensure the health workforce is equipped with the right tools to leverage data-driven technologies.
Cross-border co‑operation and partnerships are proving essential in supporting co‑ordinated and harmonised development of AI across the European Union. Strong alignment to EU initiatives, such as the 1+MG project or Genomic Data Infrastructure (GDI) initiative, is accelerating these efforts. The forthcoming roll‑out of the EHDS – which came into force in March 2025 – is also prompting a unified approach to develop AI, providing clear direction and incentives for harmonised integration.
In preparation for the EHDS, several EU Member States have begun to establish health data authorities. These entities will oversee data governance, co‑ordinate AI development and ensure that healthcare data can be securely accessed and used across sectors. They maintain compliance with EU data protection standards, support AI innovation and create a regulatory environment conducive to research and healthcare improvement.
Finland has taken a proactive approach with its Social and Health Data Permit Authority (Findata). It centralises data access services, providing a streamlined process for researchers and ensures that secondary use of health data complies with EU data privacy standards. Similarly, the Netherlands has launched the Health Data Access Body Netherlands (HDAB-NL), designed to co‑ordinate secondary data use for research, innovation and healthcare policy.
In Portugal, HealthData@PT is setting up an HDAB that will manage data governance and facilitate national and cross-border access to health data, essential for AI model development and health research. Austria is similarly developing an HDAB within its national health data infrastructure, supporting the EHDS objectives to break down data silos and enable the collaborative use of health data across the healthcare system. The Slovenian National Institute of Public Health has launched the Supporting Health Data Access Bodies in Slovenia (SI-SUD) to develop data management, metadata cataloguing, secure processing capabilities and quality labelling. All these will help researchers access health data in compliance with EHDS requirements.
Through these health data authorities, EU Member States aim to centralise the management of health data access, ensuring that AI-driven research can proceed securely and responsibly. These bodies represent a foundational element of the EHDS, supporting harmonised and efficient data sharing across borders to foster a unified AI development within healthcare. With the upcoming roll-out of the EHDS, these health data authorities will play a central role in aligning national strategies with EU goals, creating an ecosystem where AI can be leveraged for healthcare effectively.
Sources: NIJZ (2023[44]), “Supporting Health Data Access Bodies in Slovenia – SI-SUD”, https://nijz.si/nijz/javni-pozivi/javni-poziv-za-izbor-zunanjega-strokovnjaka-ki-bo-sodeloval-pri-izvedbi-evalvacije-projekta-v-okviru-pogodbe-supporting-health-data-access-bodies-in-slovenia-si-sud/; Dutch Ministry of Health, Welfare and Sport (2024[45]), “Health Data Access Body (HDAB)”, https://www.datavoorgezondheid.nl/health-data-acces-body; Findata (2024[46]), “Homepage”, https://findata.fi/en/); SPMS (2024[47]), “HealthData@PT”, Phttps://www.spms.min-saude.pt/healthdatapt-eng/about/; BKA (2024[48]), Änderung des Allgemeinen Sozialversicherungsgesetzes (BGBl. I Nr. 190/2023), https://ris.bka.gv.at/eli/bgbl/I/2023/190; EU (2025[49]), Regulation (EU) 2025/327 of the European Parliament and of the Council of 11 February 2025 on the European Health Data Space and amending Directive 2011/24/EU and Regulation (EU) 2024/2847, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ%3AL_202500327.
Most EU Member States are building a robust health data infrastructure, ensuring interoperability and expanding secondary use of health data
EU Member States are prioritising establishment of health data infrastructures by developing secure platforms to standardise, store and facilitate use of health data for primary and secondary use. The EHDS regulation provides for establishment of health data access bodies (HDABs) to grant access to health data for secondary purposes and ensure privacy and security guardrails (EU, 2025[49]). Twelve EU Member States have already established an HDAB, while ten more are doing so. However, five EU Member States have not yet made significant progress in that area. This indicates the need for further action to align with the EHDS until March 2027 – when HDABs are expected to be in place.
With the enactment of the Agreement Implementation Act 2024 and its amendment to the Health Target Steering Act, Austria established a national legal framework that enables key actors in the health system – the federal government, social insurance institutions and the federal states – to analyse data for public health. To this end, these stakeholders are jointly developing a national health data infrastructure that will serve the use of health data for public health purposes, support efficient responses to health emergencies and strengthen health system performance. At the same time, the platform lays the groundwork for key components of the national EHDS infrastructure.
Germany is laying the groundwork for the EHDS to make health data more available and useable (BMG, 2024[50]). The forthcoming Registry Act will also create a legal framework for medical registries to improve data usage for healthcare-related research, quality assurance and patient safety through regulations on facilitated data processing and linkage. In addition, it aims to improve transparency through a central hub for registry oversight (Deutscher Ärzteverlag, 2023[51]). Taken together, these initiatives aim at improving accessibility and linkability of health data across Germany and the European Union via the EHDS.
In Portugal, the OurHealth@PT and PATHeD initiatives (an EU-wide project involving 11 Member States) aim to improve secure access to health data, particularly for primary use, by integrating with the broader MyHealth@EU infrastructure. OurHealth@PT supports expansion of cross-border services, including the “Laboratory Results Reports” service, which will help share electronic health data for primary use across EU Member States. Similarly, PATHeD, which concluded in 2024, was designed to enable citizens to access their Patient Summary via a mobile app. In this way, they could present it to healthcare professionals in English or another language, further supporting cross-border care through MyHealth@EU. In parallel, Portugal is participating in the HealthData@PT joint actions, which focuses on the secondary use of health data. It supports the technical implementation of the national infrastructure, network and key services required to ensure secure access for research, innovation and health policymaking. The private sector plays a pivotal role in investing in health data platforms through secure infrastructure for managing and sharing electronic health records (EHRs), telehealth and remote monitoring services (SPMS, 2023[52]; EC, 2024[53]).
The roadmap for a national digital infrastructure in Sweden aims to centralise healthcare and dental data. In so doing, it will support efficient patient care while boosting data-driven healthcare research to advance precision medicine and training AI models (E-hälsomyndigheten, 2024[54]).
Note: Timeline from 2021 until June 2024.
Source: Data reported by EU Member States through the survey and interviews.
Achieving seamless health data exchange across the European Union relies on interoperability frameworks that enable consistent and secure communication between systems
In the Netherlands, the Electronic Health Information Exchange Act mandates electronic health data sharing among healthcare providers and supports efficient health data transfer, aligning with future EHDS provisions (Dutch Ministry of Health, Welfare and Sport, 2024[55]). Lithuania participates in the EU e‑Health Digital Service Infrastructure, which facilitates the international exchange of electronic prescriptions (e‑prescriptions) and patient summaries, enhancing cross-border healthcare through oversight of the national Health Contact Centre (LKNC) (Esveikata.lt, 2023[56]). In Portugal, interoperability initiatives include the MyHealth@EU project, which expands services like Patient Summary, e‑prescription and e‑dispensation for healthcare professionals across Europe. Meanwhile, the National Health Interoperability Platform connects public, private and social healthcare providers, allowing real-time communication across systems (EIT Health, 2024[57]).
Cyprus is developing the Xt-EHR project, a platform that defines technical standards for interoperable EHRs, critical in improving efficiency of the health data ecosystem (Xt-EHR, 2024[58]). The e‑health programme in Greece facilitates cross-border data sharing with a health hub that supports e‑prescriptions, making it easier for citizens to access healthcare across EU Member States (iED, 2024[59]). In Germany, the Federal Ministry of Health has prioritised interoperability. The Digital Act, for example, expanded the Co-ordination Office for Interoperability into a Competence Centre for Interoperability. This centre will lead work on technical and semantic standards to ensure consistent and secure health data exchange (gematik GmbH, 2024[60]).
While the national interoperability initiatives above represent meaningful progress, the overall approach across the European Union remains uneven. Continued alignment with common frameworks like the EHDS, along with improved cross-border collaboration, will be essential to achieve an interoperable European health data ecosystem. Sweden has tasked its e‑Health Agency with leading interoperability initiatives. It will work alongside the National Board of Health and Welfare to ensure a cohesive, interoperable health data infrastructure with a primary focus on semantic standards (Swedish eHealth Agency, 2024[61]).
Many EU Member States are enabling the secondary use of health data to support research, innovation and policymaking, with emphasis on AI development
Alongside other EU Member States, Portugal participates in several EU-funded initiatives, such as EUCAIM and QUANTUM, which focus on cancer imaging and data quality standards, providing valuable resources for AI-based research and clinical applications (EIBIR, 2024[62]; QUANTUM, 2024[63]). Spain has developed its National Health Data Space as part of the Digital Health Strategy, creating a national infrastructure for research and policy design (Datos.gob.es, 2024[64]). Spain also leverages platforms like HealthData 29, which makes open health datasets available to researchers, further driving health research and AI innovation (Foundation 29, 2024[65]). In Finland, the FinHITS project builds on the national infrastructure to enable international research collaboration and AI-driven analytics. It leverages the Findata platform, a legally backed service that offers secure and facilitated data access for research and policymaking (Findata, 2024[46]; Findata, 2025[66]). In France, the Health Data Hub simplifies health data access for researchers, supporting AI and data analytics projects to improve healthcare services (Health Data Hub, 2024[67]). MedReSyst and Translate-AD in Belgium foster data-sharing ecosystems to advance research into neurodegenerative conditions, such as Alzheimer’s disease (ICT & Health, 2024[68]; Stratégie de Spécialisation Intelligente Wallonne, 2024[69]).
Croatia has launched AI4Health.Cro to establish a framework for AI innovation in healthcare, enabling structured data access for research and development (R&D) (AI4Health.Cro, 2021[70]). FinHITS in Finland strengthens the country’s health data infrastructure, enabling secondary use of data through streamlined access and international collaborations (Findata, 2025[66]). In Denmark, the Research Health Data Gateway initiative is developing a centralised metadata catalogue to streamline health data access for research. This will allow researchers to integrate data from multiple sources more efficiently and accelerate AI development (Ministry of Foreign Affairs of Denmark, 2022[71]). Germany is likewise emphasising AI‑driven healthcare innovation, with 38 AI-focused projects funded from 2020 to 2025 supporting development and adoption of digital health applications (Lantzsch et al., 2022[72]). In addition, through the Medical Informatics Initiative, Germany provides clinical care data for research in a secure and accessible format. The Health Data Lab at the Federal Institute for Drugs and Medical Devices makes pseudonymised billing data from people insured in the statutory health system available for research to improve healthcare for all. Sweden is advancing secure access to health data for research, fostering AI‑driven innovation and enabling precision medicine through the AIDA Data Hub, a national e‑infrastructure designed to support data-driven healthcare research with an emphasis on medical imaging AI (AIDA, 2024[73]).
With varying maturity levels of health data ecosystems, EU Member States have begun integrating and scaling AI systems for specific use-cases in healthcare
Based on the maturity of their data ecosystem, EU Member States are integrating and scaling AI to various degrees, focusing on specific use-cases such as predictive analytics to promote AI‑driven initiatives (Box 5.2). Countries that are exploring a diverse range of AI use-cases are working to build the infrastructure and capabilities to scale these applications for broader benefits (Table 5.4).
Given the potential of AI in clinical decision support and diagnostics, countries such Greece, Lithuania, Romania and Sweden are leveraging AI to provide healthcare providers with more accurate medical imaging. Lithuania’s Oxipit initiative focuses on radiology automation for diagnostic imaging (Oxipit, 2024[74]). Meanwhile, EUCanlmage enables the linkability of clinical and genomic data to enhance cancer diagnosis through AI and precision medicine, particularly for breast, liver and colon cancers (EuCanImage, 2024[75]). The CellaVision system, developed in Sweden, uses AI to classify blood cells and support haematology diagnostics by improving accuracy in detecting abnormalities (CellaVision, 2024[76]). The start-up Advantis Medical Imaging in Greece employs AI to analyse brain magnetic resonance imaging, facilitating early diagnosis in neurology (Advantis Medical Imaging, 2024[77]). Some Romanian private healthcare facilities are starting to use Lunit INSIGHT MMG, an AI-based diagnostic tool, for detecting breast cancer on mammograms (Regina Maria, 2023[78]).
Source: Data reported by EU Member States through the survey and interviews.
As AI can improve administrative workflows, three EU Member States have launched initiatives with those specific AI applications. Poland is developing an AI-driven project to automate clinical text codification with International Classification of Diseases 10th Revision codes to improve overall efficiency and accuracy of medical documentation. Sweden is using AI-based transcription systems to convert speech to text in nearly real time, reducing the administrative burden on healthcare professionals (AI Sweden, 2024[79]). With its AI platforms, Corti (2024[80]) and Radiobotics (2024[81]), Denmark is optimising workflows spanning across administration to predictive diagnostics in radiology, which in turn helps reduce human errors.
AI-driven predictive analytics is increasingly playing an essential role in healthcare, supporting healthcare providers in forecasting health outcomes and managing chronic conditions.
In Portugal, initiatives such as Identification and Forecasting of Hospital Emergency Demand and FRAILCARE.AI leverage AI to predict peak times in emergency care and assess the needs of older patients. In this way, they enable healthcare systems to allocate resources more efficiently and tailor healthcare strategies for vulnerable populations. Similarly, the Slovak Republic uses predictive analytics to anticipate disease progression, re-admission risks and potential complications. This allows providers to make proactive, data-informed decisions that improve patient outcomes and streamline treatment pathways. In Malta, the Pharmacy of Your Choice platform uses predictive analysis to guide pharmacists towards safer, more cost-effective choices, ultimately enhancing resource management and promoting patient safety.
Despite being localised, these initiatives exemplify how AI-driven predictive analytics can integrate into healthcare systems to optimise treatments and resource allocation, contributing to a more responsive, efficient and patient-centred approach to care across the European Union.
Sources: EC (2021[82]), “Slovakia’s Recovery and Resilience Plan”, https://commission.europa.eu/business-economy-euro/economic-recovery/recovery-and-resilience-facility/country-pages/slovakias-recovery-and-resilience-plan_en?locale=en; Government of Malta (2023[83]), “The Pharmacy of Your Choice National Scheme”, https://poyc.gov.mt/en/poyc-scheme/the-pharmacy-of-your-choice-national-scheme/; FrailCare.AI (2024[84]), “Homepage”, https://frailcareai.vohcolab.org/; WEF (2024[85]), “AI in healthcare: Buckle up for change, but read this before takeoff”, https://www.weforum.org/stories/2024/01/ai-in-healthcare-buckle-up-for-big-change-but-read-this-before-takeoff/; Government of Malta (2024[86]), “Pharmacy of Your Choice”, https://health.gov.mt/poyc/.
In light of ageing populations affecting all EU Member States equally, Austria, Belgium and Croatia have initiated AI-driven projects that support home care solutions for older citizens, addressing challenges such as social isolation, safety and personalised health monitoring. The Smart and Social Home Care in Belgium aims to assist older people in private households by integrating a network of caregivers and using AI for safety and social engagement (Sirris, 2021[87]). In addition, the KALTAZARD initiative further enhances care by predicting falls and other health-related risks through sensor-based monitoring. Through the AI4Health initiative, Croatia is developing a framework for AI-driven home care health solutions. This would particularly benefit older populations in need of personalised support (AI4Health.Cro, 2021[70]). The Linked Care initiative, developed in Austria, aims to connect various health and social data source to improve services for older adults and home care patients (FFG, 2024[88]).
As far as enhancing emergency response and public health surveillance is concerned, two EU Member States are integrating AI into specific use-cases. Greece is developing a special Meteo Operational Unit. It uses machine learning to understand the spread of Coronavirus disease 2019 (COVID-19) by correlating virus variables with meteorological data for informed public health responses (NOA, 2024[89]). For its part, the Latvian Children’s Clinical University Hospital is integrating AI to support treatment management and operational efficiency (BKUS, 2024[90]).
Few EU Member States are taking steps to assess trust, understanding and acceptance of AI among the public and healthcare professionals
Across the European Union, assessing trust, understanding and acceptance of AI among public and healthcare professionals has not yet emerged as a priority. Only four EU Member States are actively pursuing initiatives in this area, underscoring the critical gap in understanding end users’ acceptance of AI.
Slovenia demonstrates a leading practice by integrating feedback mechanisms with both healthcare professionals and the public to ensure that digital health initiatives are aligned with local needs (Box 5.3). The AI4Health initiative in Croatia, funded jointly by the Digital Europe Programme and NextGenerationEU, is investing in awareness campaigns. These target healthcare professionals, promoting knowledge sharing of best practices in digital health technologies, including AI (AI4Health.Cro, 2021[70]). In the Slovak Republic, the government collaborates with medical and healthcare professional associations to promote adoption of digital health technologies, recognising the pivotal role of these associations in advocating for technological advancements. In addition, it prioritises engagement of health professionals in the co-implementation of digital health initiatives, supported by informal feedback platforms that encourage open discussions on the digitalisation of the health system. This ensures the incorporation of health workforce perceptions to foster ownership and trust of these technologies (AmCham Slovakia, 2024[91]). In Portugal, under its National Recovery and Resilience Plan (NRRP), the National Strategy for Digital Transformation in Health aims to improve equitable access to health information through digital services, such as the SNS24 portal. This platform provides citizens with streamlined access to reliable health information and digital health services (Portuguese Ministry of Health, 2024[92]).
Under the European Commission’s Structural Reform Support Programme, Slovenia developed an e‑Health strategy to advance digitalisation of its healthcare system, including integration of AI. A core priority is fostering a collaborative, patient-centred ecosystem by actively involving healthcare professionals and patients in the digital health transformation.
The strategy introduces a hybrid governance model that balances central oversight with local flexibility, allowing healthcare professionals to contribute meaningfully to e‑health initiatives and tailor solutions to local needs. Central to this model is establishment of the New Central Unit (CNU), which will standardise and co‑ordinate e‑health practices while enabling providers to adapt solutions to local needs. Advisory committees and competency units, comprising healthcare and information technology professionals, will support this effort by providing policy recommendations and operational guidance to ensure digital health solutions are both practical and aligned with clinical workflows.
To build trust and engagement among patients, Slovenia is planning regular patient surveys to continuously refine e‑health services based on user feedback. In addition, public awareness campaigns will inform the population about the benefits of digital health and improve digital literacy, fostering greater acceptance and understanding of digital health tools.
Sources: Slovenian Ministry of Health (2022[93]), Slovenija – e-zdravje za bolj zdravo družbo, https://www.gov.si/assets/ministrstva/MZ/DOKUMENTI/staro/Slovenija-E-zdravje-za-bolj-zdravo-druzbo-v2.pdf; Slovenian Ministry of Health (2024[94]), “Supporting the Digital Transformation of Healthcare”, https://www.gov.si/zbirke/projekti-in-programi/podpora-digitalni-transformaciji-zdravstva/.
Several EU Member States are conducting upskilling initiatives for the health workforce, yet EU-wide quality indicators for continued medical education are lacking
Recognising the importance of equipping healthcare professionals with the right skills to leverage AI applications, EU Member States are investing significantly in various upskilling initiatives. Four countries have included workforce upskilling in their broader digital health strategies. Meanwhile, five EU Member States have implemented open-source platforms to provide continuous digital skills training for healthcare professionals.
Targeted digital literacy and long-term learning
Latvia, the Slovak Republic and Sweden are developing targeted digital literacy programmes and long‑term AI learning courses. For their part, Belgium, Czechia and Portugal are formally integrating digital health curricula into public medical education to increase healthcare professionals’ ability to use data‑driven technologies (Box 5.4).
The Digital Health Acceleration Strategy (French Government, 2021[95]) in France and the Digital Health Strategy in Spain (Government of Spain, 2021[96]) exemplify national strategies that focus on extensive resource allocations. Both aim to ensure that healthcare professionals develop practical competencies in digital health, including AI applications.
Similarly, in Romania, Health Program 2021‑2027 aims to develop digital and AI competencies among medical personnel. It emphasises e‑health literacy as a necessary skill for remote consultations and public health data management (Government of Romania, 2023[97]). Meanwhile, the National e-Health Strategy 2030 in Bulgaria prioritises a structured approach to digitalising its healthcare system. This includes developing an organisational model for e-health and capacity building among healthcare specialists. To that end, it encourages continuous professional development to ensure effective implementation of e‑health initiatives (Ministry of Health of the Republic of Bulgaria, 2024[98]).
AI-driven training programmes
Austria, Cyprus, Ireland, Lithuania, the Netherlands and Portugal have dedicated platforms, academies and seminars offering structured data and AI-driven training programmes for healthcare professionals and the public.
The Xplain AI application is an educational tool in Austria to foster AI‑based skills in healthcare (AK Wien, 2024[99]). As part of the MyHealth@EU project, Cyprus launched the e-Health Cross-Border Health Services, which provides seminars to doctors and pharmacists to familiarise themselves with cross-border medical data exchange (NEHA, 2024[100]). For its part, in collaboration with the National Association of Pharmacies (ANF), Portugal held an informational webinar for its pharmacists on the cross-border ePrescription and eDispensation services. Two additional webinars are planned to further support professional engagement with MyHealth@EU services (SPMS, 2023[101]).
Within the Further Education and Training Strategy 2020-2024, the SOLAS initiative in Ireland aims to enhance digital literacy among the public and healthcare professionals. It offers e‑learning courses, with ongoing discussion to expand the provision of digital health literacy modules in academia (SOLAS, 2024[102]). In Lithuania, the Competency Platform for Healthcare Professionals monitors and manages the skills development of healthcare professionals, including specialists (EC, 2022[103]). Meanwhile, the Digizo.nu platform in the Netherlands supports the redesign of healthcare processes to improve efficiency through digital and hybrid integrations (Digizo.nu, 2024[104]). In addition, the Digital Skills in Healthcare initiative provides a national, open-source knowledge and information hub, offering a repository of online learning resources to enhance healthcare professionals’ digital skills and boost the digital transformation of healthcare institutions (EU, 2021[105]).
Digital literacy and long-term learning for healthcare workers
Latvia, the Slovak Republic and Sweden focus on digital literacy and long-term learning for healthcare workers, aiming to gradually build digital and AI capabilities. Vision for eHealth 2025 (Swedish Government, 2020[106]) in Sweden and Riga TechGirls programme in Latvia (Riga TechGirls, 2024[107]) encourage healthcare professionals to pursue continuous education in AI through workshops, online courses, professional development sessions and knowledge banks (Kunskapsguiden, 2024[108]). These programmes, often locally administered, emphasise accessibility and adaptability to various healthcare roles. The National Health Information Centre in the Slovak Republic extends similar AI training, equipping healthcare workers with essential skills for telemedicine, EHRs and digital data management.
In several EU countries, AI upskilling initiatives are part of EU-funded programmes and cross-border collaborations such as the HelloAI programme (HelloAI, 2024[109]) and the Digital Skills for Healthcare Transformation project (DS4Health, 2024[110]). Co-financed by the European Commission, these initiatives offer health professionals accessible online courses to apply AI in healthcare settings. Advanced master’s programmes in digital health focus on improving understanding of healthcare professionals of the design, use and development of digital health technologies.
As e-health upskilling programmes are still in the early stages of integration into medical curricula – either at the national or EU level – EU Member States have not yet focused on developing a commonly shared and EU-wide transferable skillset. Such a framework, supported by cross-country comparable indicators, could provide a consistent approach to assessing and improving continuing medical education across EU Member States.
By expanding public medical curricula in specific use-cases, three EU Member States are fostering a culture of a competitive, EU-wide health workforce. To that end, they provide the necessary skill sets to integrate data and digital technologies into clinical practice. This approach cultivates a digitally proficient and resilient workforce, actively involved healthcare providers in co-developing and co-implementing AI‑driven tools within health systems.
In Belgium, the Flanders AI Academy offers specialised courses tailored to healthcare professionals. It provides training on AI-driven diagnostics, patient monitoring and clinical decision support. In this way, the academy aims to improve clinicians’ qualifications and readiness to apply AI effectively in their workflow.
Similarly, in Portugal, the Shared Services of the Ministry of Health, part of the national public health sector, offer targeted digital health training across various healthcare roles, including doctors and nurses. This programme equips healthcare providers with basic digital skills to leverage data in different health information systems, digital health applications and platforms, cultivating a forward-oriented workforce aligned with future digital needs.
In Czechia, medical universities and the Institute for Postgraduate Medical Education, established by the Ministry of Health, have integrated digital health competencies into medical training. This integration prepares the future health workforce to work confidently with data-driven technologies.
Sources: VAIA (2024[111]), “Flanders AI Academy”, https://www.vaia.be/en/; SNS (2024[112]), “Academy SPMS”, https://academia.spms.min-saude.pt/.
Cross-border collaboration is enabling EU Member States to align their regulatory frameworks and fostering knowledge sharing of AI in healthcare
Cross-border collaboration in healthcare spans across a range of initiatives, each tailored to specific objectives, reflecting the diverse approach undertaken by EU Member States (Figure 5.3). Most EU Member States participate in the 1+MG initiative. This focuses on building secure genomic data-sharing architectures, essential for advancing research and AI-based healthcare solutions. More than half of EU Member States are actively engaged in EU-HIP, the EU health interoperability project. This initiative aims to enhance interoperability standards, aligning with platforms such as the Health Emergency Preparedness and Response Authorities (HERA) for AI-driven public health applications. In the Nordic region of Europe, five countries are collaborating to establish common technical and semantic frameworks to support cross-border e-health efforts (Box 5.5).
To strengthen international collaboration, the Global Digital Health Partnership (GDHP), with the participation of 14 EU Member States, fosters development of shared data standards and frameworks to support interoperability and digital health innovation. Research-driven collaborations, such as the ELIXIR project, are widely supported by 21 EU Member States to foster data sharing for genomics and personalised medicine. Similarly, the Transforming Health and Care Systems (THCS) project, involving 21 European countries, aims to create a framework for healthcare transformation, including AI testing in clinical settings.
In addition, Denmark, Germany and Portugal participate in the International Medical Device Regulators Forum (IMDRF). They focus on regulating AI-enabled medical devices to ensure safety and efficacy. Initiatives on AI governance and ethics, led by the World Health Organization (WHO) and the Organisation for Economic Co‑operation and Development (OECD), include contributions from various countries, promoting global standards and responsible AI adoption in healthcare.
Austria, Denmark, the Netherlands, Portugal, Slovenia and Sweden are investing in cross-border collaboration that focuses on harmonising healthcare standards and regulatory frameworks. These efforts aim to align AI regulations within broader digital health frameworks. The THCS partnership, involving 21 EU Member States, aims to create a framework for healthcare transformation, including AI testing in clinical settings (THCS, 2022[113]). Similarly, the IMDRF, with the participation of Denmark, Germany and Portugal, has a working group on developing standards for AI-enabled medical devices, prioritising patient safety and performance requirements (IMDRF, 2024[114]).
Collaboration revolving around research and innovation (R&I), involving 24 EU Member States, supports the 1+MG initiative to build a secure data-sharing architecture for genomics, essential for AI-based research (GDI, 2024[115]).
Source: Data reported by EU Member States through the survey and interviews.
Czechia, Portugal and the Netherlands reported initiatives that focus on enhancing data-sharing standards for AI. The participation of Portugal in EU-HIP, alongside 15 other EU Member States, aims to ensure interoperability with the HERA platform, facilitating AI applications in public health (Statens Serum Institut, 2025[116]). Similarly, the GDHP, with the participation of 14 EU Member States, promotes international collaboration to develop and implement shared data standards and frameworks to support interoperability (GDHP, 2024[117]).
Several EU Member States are participating in international collaborations focused on sharing best practices for AI-driven healthcare. Supported by the WHO and the OECD, these partnerships provide platforms for countries to share knowledge on AI governance, ethics and real-world AI applications in healthcare. Through these collaborations, countries contribute to global knowledge-sharing practices that promote the responsible use of AI in healthcare by setting standards and guidelines.
Nordic countries, including Denmark, Finland, Iceland, Norway and Sweden, are collaboratively advancing cross-border e‑health initiatives under the guidance of the Nordic Council of Ministers.
A central focus of the Nordic collaboration is the establishment of standardised electronic health records with an emphasis on technical and semantic harmonisation across countries. This alignment enables the secure, efficient exchange of health information across borders, ensuring that patients can access their medical records and receive co‑ordinated care throughout the Nordic region.
The collaboration extends to establishing common frameworks for mobile health applications, health databases and registries that enable secondary health data use for research, public health and policy development. By creating shared platforms and databases, the initiative facilitates high-quality data standards and innovation in healthcare, giving researchers and policymakers access to comprehensive cross-border health data. This approach strengthens public health monitoring and provides deeper insights into health trends across the Nordic countries, enhancing both preparedness and responsiveness in public health.
A notable feature of this partnership is its focus on clinician support. By implementing e‑health standards that streamline documentation and reduce administrative tasks, the Reducing Clinician Burden initiative aims to free up clinicians’ time, allowing them to focus on delivering quality care and enhancing patient outcomes.
Sources: Nordic Council of Ministers (2019[118]), e-Health Standardisation in the Nordic Countries, https://norden.diva-portal.org/smash/get/diva2:1340369/FULLTEXT01.pdf; Nordic Council of Ministers (2024[119]), “Supporting the Healthcare Professionals’ Work and Data Quality through e-Health Standards”, https://pub.norden.org/temanord2024-514/index.html.
Most EU Member States are investing significant resources in cross-border, research-driven genomic initiatives, while a smaller subset is pursuing policy-led strategies or developing dedicated genomic infrastructure
The alignment of countries with the 1+MG initiative reflects a range of focal areas and objectives. Most EU Member States focus on research-driven collaboration, like the pan-European ELIXIR project, the Danish National Genome Center and Genomic Medicine Sweden (GMS), to facilitate data sharing and personalised medicine development.
Approximately one‑third of EU Member States, including Belgium, Bulgaria, Czechia, Germany, Lithuania and Sweden, have launched dedicated national genome projects or established genomic infrastructures, often linked to biobanks or genomic databases. Cyprus, for instance, has established its first national biobank, which is integrated into the Genome of Europe (GoE) network and contributes to broader European genomic research efforts (Box 5.6). The European GDI project, with participation from 24 Member States, further supports the creation of EU-wide genomic infrastructure. Meanwhile, less than 10% of countries, such as Finland, Ireland and Portugal, are advancing their initiatives through policy-focused strategy led by national co‑ordination bodies or integrated into broader health strategies (Figure 5.4).
Note: Belgium, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Luxembourg, Malta, the Netherlands, Norway, Portugal, Romania, Slovenia, Spain and Sweden are advancing through research-driven collaboration; Belgium, Bulgaria, Cyprus, Czechia, Germany, Latvia, Lithuania, Slovak Republic, Slovenia and Sweden are advancing with dedicated national genome projects and/or established genomic infrastructures; Finland, Ireland and Portugal are advancing through policy-focused strategies and/or integration into broader health strategies.
Source: Data reported by EU Member States through the survey and interviews.
Belgium, Bulgaria, Cyprus, Czechia, Finland, France, Germany, Latvia, Lithuania, the Slovak Republic, Slovenia and Romania have launched dedicated national genome projects or established specific genomic infrastructures, often linked to the development of biobanks and genomic databases.
In Germany, the genomDE National Strategy for Genomic Medicine aims to establish a robust data infrastructure, while addressing ethical, regulatory and safety considerations to ensure its proper and responsible implementation (BMG, 2024[120]). Lithuania is developing a Lithuanian genome project through establishment of a national GDI (Minister of Health of the Republic of Lithuania, 2023[121]). The partnership involves the Vilnius University Hospital Santaros Clinics, the Lithuanian University of Health Sciences Hospital Kaunas Clinics and the National Cancer Institute. Latvia is building a national genome reference and associated IT infrastructure, strengthening its capacity for genomic research and data sharing within the GDI framework (Latvian Biomedical Research and Study Center, 2020[122]).
Other countries, such as Finland, Ireland and Portugal, are advancing the initiative through policy‑focused strategies and co‑ordination by national bodies. In Finland, a national co‑ordination group under the guidance of the Ministry of Social Affairs and Health is directing efforts (STM, 2024[123]). Meanwhile, Portugal has incorporated 1+MG objectives into its National Strategy for Genomic Medicine (Government of Ireland, 2021[124]).
Research-focused collaborations also play a pivotal role in advancing the 1+MG initiative. A network of 21 EU Member States is working on ELIXIR (2024[125]), a pan-European infrastructure supporting genomic data sharing. In addition, since 2018, Sweden has been developing GMS through a national platform to ensure patients have equal and cost-effective access to genetics analysis (GMS, 2024[126]). Similarly, the Danish National Genome Center was created to improve diagnostics and personalised treatment through whole-genome sequencing (NGC, 2024[127]).
In addition, 24 EU Member States are investing in federated genomic data sharing through the GDI. This initiative aims to create secure, interoperable frameworks for cross-border data exchange. In this way, it will facilitate collaborative research and advance genomic medicine across the European Union (GDI, 2024[115]).
Biobank.cy, the first national biobank in Cyprus, serves as a centralised repository for Cypriot genomic data, ensuring that data are collected, securely stored and made accessible while following ethical standards. By housing genetic samples of over 1% of the population in Cyprus, Biobank.cy can support a wide range of research applications – from genetic disease studies to personalised medicine approaches.
The biobank’s integration with the Genome of Europe (GoE) network – spanning 26 EU Member States and aligned to the 1+MG initiative – will enhance its strategic relevance. In this way, it will enable Cyprus to contribute to a broader EU database, facilitating comparative genomic studies and cross-border research collaboration. Biobank.cy has prioritised adherence to the latest data-sharing protocols and interoperability standards, ensuring that data can be shared securely and effectively under the EHDS framework.
The example of Biobank.cy demonstrates how smaller countries can create robust genomic infrastructure that supports large-scale data integration and innovation. By contributing to the GoE network, Cyprus provides diversity to the EU genomic dataset. This is crucial for developing treatments that account for the genetic variability across EU populations.
Source: CY-Biobank (2024[128]), “The Center”, https://biobank.cy/profile/.
Few EU Member States rely on EU financial mechanisms such as the RRF to advance AI in health, highlighting uneven national priorities
Despite eligibility across all 27 EU Member States, only 7 use RRF funds to advance AI in healthcare. Bulgaria, Ireland, Latvia, Lithuania, Portugal, the Slovak Republic and Slovenia report investments targeting AI advancements across various health areas (Figure 5.5). Portugal stands out with its Health Data Lake initiative, which centralises health data into a secure, interoperable platform aligned with EHDS standards (Box 5.7). The limited uptake suggests that these seven EU Member States have prioritised AI within their healthcare investment strategies. Meanwhile, others may have chosen not to allocate RRF funds to this area, potentially due to differing priorities or lack of immediate need to draw on these resources.
Source: Data reported by EU Member States through the survey and interviews.
Ireland and Slovenia are integrating AI into their e-health systems, focusing on projects that improve data accessibility, patient management and healthcare delivery. In Ireland, the NRRP directs RRF funds towards community e‑health, e‑pharmacy and digital financial management to modernise healthcare services (Government of Ireland, 2021[124]). Slovenia, with EUR 101 million in RRF funds, is expanding telemedicine, digitalising medical records and developing AI-driven digital services (EC, 2021[82]).
Aligned with the 1+MG initiative, Latvia and Lithuania are investing RRF funds to build national genomic databases and IT infrastructures. This is intended to enable genomic data collection, storage and analysis, enhancing each EU Member State’s contribution to the GDI (Latvian Biomedical Research and Study Center, 2020[122]; Minister of Health of the Republic of Lithuania, 2023[121]).
RRF funds are also being used to develop AI healthcare applications. Bulgaria is investing around EUR 14 million into a National Digital Platform for Medical Diagnostics. This applies AI for real-time health data processing, supporting diagnostic accuracy and remote monitoring to improve patient outcomes (EC, 2024[129]).
Another aspect of RRF funding is its allocation towards workforce training and upskilling in digital health competencies. The Slovak Republic, for instance, includes workforce upskilling in digital health as part of its AI integration efforts, ensuring that healthcare professionals can use and manage new AI tools effectively (EC, 2021[82]).
AI is playing an increasingly central role in transforming the public sector. AI technologies enable governments to significantly improve operational efficiency by automating routine tasks, optimising workflows and reducing administrative processes. Natural language processing, machine learning and computer vision applications are helping public agencies sort citizen requests, automate eligibility checks for services and provide 24-hour/7-days-a-week digital assistance (OECD, 2025[130]).
In Portugal, Health Data Lake is a leading initiative to centralise and securely manage health data from public and private sources. This platform, supported by a EUR 10.5 million of RRF investment, is designed to consolidate EHRs, medical imaging and genomic data into a unified, interoperable repository. This, in turn, is aligned with EHDS standards and regulated by a national legal framework. As such, it provides a reliable foundation for training AI models on applications such as predictive analytics, diagnostics and personalised treatments. By consolidating data across multiple sources, the Health Data Lake ensures consistent standards and formats. It enhances the reliability and relevance of AI models, while fostering collaboration among healthcare providers, researchers and developers to drive innovation ethically and responsibly.
Source: EIT Health (2024[57]), European Health Data Space in Portugal, https://eithealth.eu/wp-content/uploads/2024/06/EHDS-white-paper-Portugal.pdf.
AI is playing an increasingly central role in transforming the public sector.1 AI technologies enable governments to significantly improve operational efficiency by automating routine tasks, optimising workflows and reducing administrative processes. Natural language processing, machine learning and computer vision applications are helping public agencies sort citizen requests, automate eligibility checks for services and provide 24/7 digital assistance (OECD, 2025[130]). AI also holds potential to enhance policy design and implementation, although adoption rates are lower than for its use in improving internal government operations (OECD, 2024[131]).
Moreover, AI empowers governments to deliver more responsive and personalised services that better meet citizens’ evolving needs. Public institutions can now gain deeper insights into citizen requests through sophisticated data analysis, developing proactive responses to complex challenges. AI‑powered systems are enabling the shift from one-size-fits-all approaches towards context-aware interactions tailored to individual circumstances. For example, AI chatbots have improved response times in welfare systems, and predictive analytics helps anticipate infrastructure needs and optimise budgeting decisions (OECD, 2025[130]). These innovations reflect a growing commitment to human-centred public services that prioritise accessibility, responsiveness and user empowerment.
Despite these promising developments, the deployment of AI in the public sector raises important concerns. Risks include the potential for skewed outcomes, lack of algorithmic transparency, data privacy breaches and erosion of trust in automated decision-making processes. Additionally, structural barriers continue to hinder effective and coherent AI adoption across government agencies (OECD, 2025[130]). These include fragmented data infrastructures, outdated legacy information technology (IT) systems, limited digital capabilities within administrations and inadequate procurement frameworks. Traditional procurement and talent acquisition processes may struggle to accommodate the speed and specificity of AI innovation (OECD, 2024[132]). Disparities in digital maturity across different levels and functions of government exacerbate implementation challenges.
These challenges can be addressed through measures that combine the right enablers, guardrails and engagement processes (OECD, 2025[130]). Enablers notably include robust governance structures, institutional safeguards and targeted digital capacity building. Equally important is addressing the digital skills gap through comprehensive training programmes for civil servants, recruitment of AI specialists and partnerships with educational institutions to build and attract talent. Governments should invest in digital literacy initiatives across all administrative levels, creating clear career pathways for technology specialists in the public sector. Guardrails, in turn, include transparent algorithm registries, capacity building for civil servants and oversight, engaging a broad range of stakeholders. Embedding transparency, human oversight and participatory design (another example of engagement process) across the AI policy cycle helps ensure responsible innovation in the public sector (OECD, 2024[132]).
1. OECD work has also evidenced the use of AI for improving regulatory design (OECD, 2025[130]). Moreover, data from the OECD’s Digital Government Index suggest that AI tends to be used much more often for improving internal governmental processes than for enhancing policy design and implementation (OECD, 2024[131]).
The EU Coordinated Plan on AI identifies AI as a crucial technology for enhancing public services, improving citizen-government interactions, enabling smarter analytical capabilities and increasing efficiency across public sector domains. It further recognises that implementation of AI systems can support democratic processes, bring benefits across all key public sector activities. Through early adoption, the public sector can lead the way in implementing AI that is safe, secure and trustworthy. To that end, the EU Coordinated Plan on AI encourages Member States to:
Take full advantage of the opportunities offered by RRF by including in their national recovery and resilience plans measures focusing (for example) on building capacity to seize the advantages of predictive analytics and AI in policymaking and public service delivery. The proposed reforms and investments under this component champion the RRF Flagship ‘Modernise’ focusing on digitalisation of public administration and services, including judicial and healthcare systems. They might also mirror the objectives of the RRF Flagship ‘Reskill and upskill’, by providing skills and new competences for civil servants and managers, notably in relation to green and digital transitions and to enhancing innovation in public administration.
Use of AI in the public sector is a key focus area for EU Member States
Most EU Member States (16) identified promoting AI adoption in the public sector as a priority in their national AI strategies (see Chapter 2). In total, 91 initiatives were reported across 24 EU Member States (Figure 5.6). There is significant activity in areas such as improving administrative systems and establishing governance frameworks. However, other areas – such as training programmes for civil servants – are reported less frequently.
Half of EU Member States reported initiatives to develop AI-powered administrative systems and public services, making this the most common area of interest for AI use in the public sector. These initiatives aim to automate workflows, enhance service delivery and reduce administrative burdens.
Several countries have deployed AI systems for document processing and workflow automation. Through the AI-Supported Administrative Notification project, Austria is piloting automation possibilities in administrative processes through handwriting recognition, translation and knowledge management. The Cadastre system in Greece has automated the reading and categorisation of property contracts, applying relevant legal rules to generate assessments for approval. This has significantly reduced processing time from several hours to less than 10 minutes and cut costs from EUR 15 to just EUR 0.11 per assessment. In Italy, the PRODIGIT Project collaborates between the Presidential Council of Tax Justice and the Ministry of Economy and Finance to modernise judicial tax proceedings through AI-powered tools, enhancing efficiency in tax‑related legal processes.
Source: Data reported by EU Member States through the survey and interviews.
Citizen-facing AI services represent another significant trend. In Latvia, an AI virtual assistant/chatbot network creates a centralised platform for managing government websites, improving public access to information. Malta has deployed an AI chatbot providing 24-hour/7-days-a-week assistance to users seeking information about public services. ION in Romania serves as an AI counsellor to link citizens and the executive branch, capturing sentiments of Romanian citizens in real time. Portugal has implemented multiple AI assistants, including a social security chatbot created during the COVID‑19 crisis and the Caixa Geral de Aposentações AI Virtual Assistant to enhance user navigation and reduce response wait times (CGA, 2023[133]). Poland aims to integrate the Polish Large Language Universal Model into mObywatel, a virtual assistant for citizens to use public administration services, to automate document processing, content analysis, information search and support in answering citizens' questions (GOV.pl, 2025[134]).
AI for regulatory compliance and specialised administrative tasks is gaining traction. Czechia developed an analytical AI platform to identify discrepancies between different sets of draft legislations. In Latvia, AI in the Evaluation of EU Projects assesses the applicability of generative AI for analysing procurement documentation related to EU-funded projects with the aim of testing AI prototypes for public sector applications.
Tax administration is also emerging as a key domain for AI adoption. In Latvia, the Artificial Intelligence in Preparing Responses to Taxpayer Inquiries programme explores AI applications in tax administration by using AI tools to generate responses on labour taxes. In Italy, the AI in Tax Administration leverages advanced analytics software to prevent tax avoidance by analysing data from the Tax Registry and Archive of Financial Relations. Finally, the financial administration in Slovenia uses AI in processing value-added tax returns to increase the effectiveness of controls through predictive analytics.
Countries are establishing governance frameworks and ethical guidelines for responsible AI use in the public sector
Sixteen EU Member States have developed AI governance frameworks and ethical guidelines in the public sector, reflecting growing recognition that AI adoption must be grounded in transparency, accountability and public trust. In all, 23 initiatives have been launched, with approaches ranging from central co‑ordination hubs to legislative guidance and ethical charters.
Several countries have developed comprehensive AI advisory and oversight structures.
In Austria, the AI Directory creates a comprehensive overview of AI systems in use in federal public administration, identifying challenges and risks related to AI in the public sector. A complementary AI Map serves as a cross-ministerial register of ongoing and planned AI initiatives. The country has also published a Practical Guide on Digital Administration and Ethics to help public sector servants navigate AI use, covering opportunities, challenges, ethical issues and legal frameworks.
The Netherlands has pioneered algorithmic accountability frameworks with two key initiatives. The Algorithm and AI Register catalogues systems in use by the public sector, enhancing transparency. This is complemented by Fundamental Rights and Algorithm Impact Assessment, which involved 18 pilot projects across public sector organisations. These aimed to ensure that algorithm deployment aligns with human rights principles and prevents unintended negative consequences. This systematic approach to algorithmic assessment and transparency offers a model for other EU Member States considering similar governance mechanisms.
Germany is establishing an AI Advisory Centre as a central co‑ordination hub for AI adoption across the federal public sector. With a budget of EUR 9.18 million, the centre will offer legal, technical and ethical support to government ministries and agencies. The country has also developed an AI Framework that focuses on human-centred design, transparency and fairness, and is developing AI Guidelines to ensure a harmonised inter-ministerial approach to AI use.
Several other EU Member States are developing similar ethical principles and guidelines for AI implementation. Belgium is developing a Charter for the Responsible Use of AI in Public Services to define ethical principles for AI in public services, ensuring trust and accountability. Finland has issued guidelines for responsible use of generative AI in government. In Ireland, Interim Guidelines for Use of AI in the Public Service provide operational principles for government departments, with a full framework expected in 2025. (Government of Ireland, 2021[124]). Luxembourg has adopted an AI Charter outlining ten key guidelines for responsible use of AI in parliamentary activities (Chamber of Deputies, 2024[135]). Malta, in turn, issued its Generative AI Tools Usage Policy for its public administration in September 2024. This policy seeks to foster human-centred accountability and transparency, security and safety, data confidentiality, responsible innovation and sound complaint handling practices.
Several EU Member States have adopted sector-specific or institutional approaches. In Denmark, the Digital Taskforce for Artificial Intelligence is working across agencies to remove deployment barriers and accelerate responsible scaling. AI Guidelines for Employment and Social Protection Services in Germany focus on ethical and inclusive deployment in social programmes. Sweden has tasked the Swedish Agency for Digital Government and Swedish Authority for Privacy Protection with developing rules for using generative AI in public administration. In Lithuania, the Working Group on Artificial Intelligence within the Committee for the Future of the Seimas (Parliament of the Republic of Lithuania) advises on AI policy and fosters a culture of innovation in government. Innovation hubs and experimental spaces for AI in the public sector are gaining momentum.
Twelve EU Member States have established innovation hubs and experimental initiatives to accelerate AI adoption in the public sector. These initiatives create environments where new AI technologies can be safely tested, refined and deployed to address government challenges. The 22 reported initiatives notably include sandboxes, innovation labs, marketplace solutions and collaborative platforms.
Environments for AI testing and experimentation in the public sector have developed in recent years
Several countries have created dedicated testing environments for AI experimentation under controlled conditions. This aligns with the EU Artificial Intelligence Act (AI Act) requirement that each Member State establish at least one AI regulatory sandbox by 2 August 2026, or participate in joint sandboxes with other countries. In Greece, the GovTech AI Sandbox helps public sector organisations overcome barriers to innovation by providing a supervised space to deploy and test AI solutions before wider implementation. The Slovak Republic is developing Experimental Regulatory Environments for AI with preparation scheduled for 2025, focusing on legislative analysis and methodologies aligned with the EU AI Act. Ireland has created a “safe space” sandbox where civil servants can experiment with AI tools without operational risks. Denmark has a sandbox for testing AI solutions as part of its broader plan for new technology and automation of the public sector.
Multiple countries have established dedicated platforms to accelerate AI adoption. In Germany, Datalabs recruit experts from academia, civil society and the private sector to support federal administration in implementing AI-driven solutions. The Law as Code project in Austria uses symbolic AI to implement laws and legally binding documents in machine-readable formats. In Slovenia, the Semantic Analyser applies natural language processing to analyse legal texts and identify essential concepts. Lithuania operates the GovTech Lab to help public sector institutions identify challenges addressable through AI and other emerging technologies.
Several initiatives focus on creating the technical foundations for AI experimentation. Germany has developed the AI Platform KIPITZ (2024), an overarching infrastructure supporting large language models for federal administration. It is also creating an AI Marketplace for Opportunities to connect ministries with AI solutions (BMI, 2024[136]). Smart digital public services (2024‑2029) in Slovenia establishes an interoperable ecosystem supporting algorithmic and analytical tools. In Latvia, the Memorandum of Understanding with Microsoft (2024) establishes a national centre for AI to accelerate digital transformation in public administration.
Several EU Member States have created structured frameworks for collaborative innovation. The Digital Wallonia strategy in Belgium aims to transform Wallonia into a digital platform by innovating public services and serving as a testing ground for scalable digital solutions (Digital Wallonia, 2022[137]). In Sweden, the eSam collaboration brings together 41 government agencies implementing AI initiatives, while the Collaboration for AI in Municipalities provides tailored support to local governments. The GovTech Challenge Series (EUR 1 million annually) in Lithuania connects public sector challenges with providers of innovative solutions through a structured process. This is supported by the country’s GovTech Sandbox (2024) involving 14 public organisations in experimental AI concepts. In Luxembourg, the Public Sector Innovation Hub (2020) serves as a platform for collaboration between government entities and technology innovators (Government of Luxembourg, 2024[138]).
Some countries are establishing specialised incubators for public sector AI solutions. In Spain, the incubator of AI solutions for the public administration focuses on generative AI use-cases, with more than 300 requests from ministries. In Portugal, an innovation approach applies AI to enhance regulatory impact assessments through the PLANAPP Project (PLANAPP, 2021[139]). An initiative in Belgium, A Land for Tomorrow, uses AI to process citizen proposals for government reform, generating insights for policymaking.
These diverse innovation approaches demonstrate a growing recognition that dedicated experimental spaces are essential for successful AI adoption in government. By providing structured environments where new technologies can be safely tested and refined, EU Member States are creating pathways for AI solutions to move from concept to operational deployment. At the same time, they are managing technical and organisational risks.
AI training and capacity building for civil servants remain limited in EU Member States
Eleven EU Member States have reported initiatives focused on AI training and capacity building for civil servants. These programmes aim to develop AI literacy and skills among government employees, enabling them to implement and manage AI systems effectively. The 15 reported initiatives range from basic awareness programmes to specialised technical training.
Several countries have established comprehensive AI education programmes for public officials. In Belgium, the AI Expertise Centre at the Flanders Digital Agency identifies the needs of civil servants for AI upskilling and reskilling. In so doing, it supports public sector innovation by equipping civil servants to use AI for improved policymaking and operational efficiency. The Institute of Public Administration in Bulgaria offers training (EUR 16 500) in AI applications through multiple formats, including self-learning e‑modules, face-to-face training and interactive workshops. In Latvia, the E‑course for Basic Competence Level in AI for Public Administration Employees will train 1 000 public administration employees through four modules: introduction to AI; AI concepts and applications; AI issues and ethics; and future and practical use of AI. Romania has an advanced digital skills training programme for civil servants (EUR 20 million through the RRF). It aims to train 32 500 civil servants (including 2 500 senior people) in advanced digital skills, including database management, system management, business analysis, data analysis and programming.
Specialised AI training is targeting specific competency needs. The knowledge centre for municipalities in Denmark helps share knowledge about digitalisation and explore how AI can create value in municipalities. In France, AI-related training for civil servants includes the Campus du numérique public, which covers digital topics and methods, including AI-related skills. Ireland offers capacity building for civil and public servants through one-day courses administered by the Department of Public Expenditure, Infrastructure, Public Service Reform and Digitalisation. In addition, about 120 public servants have completed a Certificate in Foundations of Artificial Intelligence (part-time over 12 weeks). Luxembourg has also developed digital learning resources focused on emerging technologies, including AI. In Malta, the Institute for the Public Services has integrated AI training into its curriculum, including a module that introduces AI.
Some initiatives focus on generative AI and its applications in government. The KURSUOK one-stop-shop platform in Lithuania is also available for civil servants, providing AI education resources. In Slovenia, the Strengthening the Digital Skills of Civil Servants programme aims to improve competencies in core digital skills, including specialised skills on AI. It was designed in accordance with the OECD Framework for Digital Talents and Skills in the Public Sector.
AI is playing an increasingly transformative role across the mobility sector, driving significant gains in efficiency, sustainability and safety; optimising transport operations; and enhancing overall mobility systems. In public transport, AI enhances scheduling, optimises multimodal mobility and improves passenger experiences through predictive analytics (Moumen, Abouchabaka and Raflia, 2023[140]). In freight and logistics, AI-driven route optimisation, demand forecasting and predictive maintenance boost supply chain efficiency and reduce operational disruptions (Du Plessis et al., 2025[141]). Automated vehicle (AV) technologies rely on AI for perception, navigation and real-time decision making, while AI‑powered traffic management systems use dynamic signal control and congestion prediction to improve road capacity and flow (Garikapati and Shetiya, 2024[142]).
More broadly, AI-driven automation streamlines transport operations, reducing delays and improving service reliability. AI-powered traffic optimisation lowers congestion, shortens travel times and decreases emissions, supporting sustainability goals (Nikitas et al., 2020[143]). Advanced driver-assistance systems and AI-based hazard detection enhance road safety, reducing accidents and improving overall transport security (Fernández-Llorca and Gómez, 2023[144]). AI also plays a role in the transition to less resource-intensive mobility, optimising electric vehicle charging networks, improving fuel efficiency in freight transport and supporting low-emission urban mobility strategies (de Queiroz et al., 2021[145]).
Despite these advantages, AI adoption in mobility faces multiple challenges. Key barriers are data accessibility and interoperability, as transport networks rely on vast datasets collected from multiple sources, including public authorities, private mobility providers and IoT devices. Fragmented data infrastructures and inconsistent data-sharing protocols can hinder the ability of AI to optimise mobility services (Paiva et al., 2021[146]). Additionally, high implementation costs and the need for advanced computational resources pose financial challenges, particularly for SMEs in the mobility sector (Ma et al., 2020[147]). Ethical and regulatory concerns, such as liability in automated driving, algorithmic biases in AI-driven traffic enforcement and cybersecurity risks in connected vehicle networks, are further barriers for large-scale AI deployment (Bathla et al., 2022[148]).
Maximising the benefits of AI in mobility
Given the transformative potential of AI in mobility, policy frameworks play a critical role in maximising benefits while addressing these challenges. Standardising data governance and promoting open-data platforms can facilitate AI adoption by ensuring interoperability and secure data-sharing mechanisms (Craglia et al., 2021[149]). Investment in AI-specific infrastructure can further support the scalability and efficiency of AI in mobility applications. This could include high-performance computing resources and 5G‑enabled transport networks (Wang et al., 2023[150]). Regulations help mitigate risks related to transparency, safety and accountability (Craglia et al., 2021[149]). By fostering innovation-friendly regulations, incentivising AI adoption and ensuring that AI-driven mobility solutions align with sustainability goals, policymakers can position AI as a key driver of the future transport ecosystem in the European Union.
The update to the plan reinforced the need for strategic investment in AI-driven mobility, highlighting the importance of data availability, interoperability and cross-border co‑operation. The goal is to maintain EU leadership in smart mobility, while ensuring that AI deployment aligns with key EU values of safety, transparency and inclusivity.
The EU Coordinated Plan on AI encourages Member States to undertake several actions related to AI in mobility:
Actively promote the development and testing of AI technologies in automated functions for all modes of transport, with the help of the relevant European partnerships.
Analyse and facilitate the deployment of trustworthy AI solutions in all modes of transport that can enhance efficiency with the help of automated mobility services and freight transport operations in order to reduce the burden on the environment.
Share lessons learned from R&I projects and pilots to create a European common knowledge base
Assess the potential of vehicle automation for urban transport and support cities in their transition while rethinking mobility systems, including public transport services, infrastructure maintenance, logistics, fares and regulation; and
Take full advantage of the opportunities offered by RRF, for instance in line with the actions described in the example of component on ‘Clean, smart and fair urban mobility’.
These actions aim to position the European Union as a global leader in AI-powered sustainable mobility, ensuring that AI adoption aligns with key EU values. By fostering collaboration between Member States, industry and research institutions, the EU Coordinated Plan on AI aims to enhance EU competitiveness in AI-driven transport, while ensuring that mobility innovations benefit all citizens.
Initiatives to foster AI adoption in mobility vary significantly across EU Member States
More than half of EU Member States (17 of 27) have launched initiatives to foster AI adoption in mobility, although their distribution remains uneven. Certain areas – such as AVs and urban mobility transformation – have received more policy attention. Conversely, data sharing for mobility appears to be in its early stages. In total, 49 AI-driven mobility initiatives have been reported.
Note: AV stands for automated vehicle.
Source: Data reported by EU Member States through the survey and interviews.
Automated vehicles are a major focus of AI in mobility
Eleven EU Member States have reported initiatives to foster development and deployment of AVs. While this signals interest in AI-driven mobility innovation, the absence of dedicated AV policies or projects in over half of EU Member States highlights uneven progress across the region. The 18 reported initiatives refer to testing and innovation hubs, regulatory adaptations, AI safety and verification, and AV integration into public transport and urban mobility.
Several countries have established dedicated AV testing environments to advance research and real-world deployment. In Austria, Automated Transport Innovation Labs provide physical and digital testing infrastructures for AVs, funded by the Ministry for Climate Action, Environment, Energy, Mobility, Innovation and Technology (BMK) (ALP.Lab, 2024[151]; Digitrans, 2024[152]). In Belgium, Mobilidata 2 Proloog focuses on real-world trials of remote driving technologies, including a Remote Driving Test Week (MobiliData, 2024[153]). Additionally, the Flemish Task Force on Automated Driving, launched in 2023, connects research institutions, stakeholders and logistics operators to create a more robust AV innovation ecosystem. Slovenia has launched the SRIP ACS+ Strategic Research Innovation Partnership, which facilitates AV testing through an industry-driven platform that integrates research institutions and manufacturers (Slovenian Ministry of Higher Education, Science and Innovation, 2023[154]). Sweden has taken a broad approach with Drive Sweden (2024[155]), a national platform involving over 200 stakeholders from business, academia and government, aimed at testing connected and automated mobility solutions.
Some governments are adapting legislative frameworks to enable AV deployment while ensuring responsible governance. Finland has initiated Legislative Amendments for Road Transport Automation, which is undergoing public consultation; a government proposal was expected by 2025 (Finnish Ministry of Transport and Communications, 2021[39]). Czechia has followed an ethical governance approach by setting up the Ethics Commission for the Assessment of Automated and Autonomous Vehicles. It provides guidelines on responsible AI use in AV research and deployment, ensuring that ethical considerations are embedded in the country’s automated mobility strategy (Czech Ministry of Transportation, 2024[156]).
Ensuring AI safety in AVs remains a key policy area. Germany has introduced several large-scale AI‑driven AV initiatives. The Safe AI Engineering project, funded with EUR 17.2 million, focuses on verifying AI systems in AVs to ensure safety and reliability (BMWK, 2023[157]). AI Applications for Connected and Automated Driving supports multiple projects exploring AI-driven automation, safety enhancements and the integration of Level 4 AVs into public transport, with an investment of EUR 47.2 million (BMDV, 2018[158]; BMDV, 2021[159]). Another initiative, nxtAIM, aims to develop and train generative AI methods for automated driving, focusing on overcoming key barriers to the deployment of automated and autonomous vehicles (nxtAIM, 2025[160]). France has invested in AI-driven AV safety through the PRISSMA project, a EUR 13 million initiative that validates AV systems with AI components. Additionally, the IRT SystemX research centre contributes to AV safety through multiple projects co-funded by public and private stakeholders. In Sweden, the Traffic-safe Automation initiative, led by the Strategic Vehicle Research and Innovation programme, seeks to position road safety as a key driver of automation. The initiative aims to enhance transport sustainability through safe, connected and automated vehicles designed for shared use (FFI, 2025[161]).
Some EU Member States are integrating AVs into public transport and urban mobility strategies. Croatia has invested significantly in AV infrastructure through Autonomous Vehicles in Zagreb, a EUR 452.7 million project, including EUR 179.5 million from the RRF. Ireland has set up the testing authorisation regime for connected and automated mobility on public roads. It also has a large-scale pilot focusing on Cooperative Intelligent Transport Systems (C‑ITS) to facilitate safe deployment of AV (Transport Infrastructure Ireland, 2025[162]). Additionally, the Smart Dublin (2022[163]) initiative is trialling AI-powered last-mile delivery robots using computer vision technologies. Belgium is advancing automated public transport solutions through NAVAJO, a pilot project introducing an automated shuttle service in Ottignies-Louvain-la-Neuve. Funded under Digital Wallonia’s Intelligent Territories programme, the project tests and validates AV technologies for integration into urban mobility frameworks (TEC, 2022[164]). Portugal has launched the Route 25 Project to develop connected and automated transport systems. The total investment of EUR 80 million includes EUR 32.5 million in RRF funding (Instituto de Telecomunicações, 2025[165]).
AI is often being integrated into broader sustainable mobility projects
Eight EU Member States have reported initiatives to integrate AI into larger programmes to foster sustainable mobility solutions, supporting the transition to resilient and efficient transport systems. These efforts focus on AI-driven optimisation of mobility systems; multimodal transport integration; innovation in mobility R&D; and long-term strategic planning for transport digitalisation.
EU Member States are investing in AI-driven optimisation of transport systems to improve efficiency, reduce emissions and enhance traffic management. Germany is funding the EUR 75 million AI Applications in Mobility and Logistics programme. It supports AI-based analysis, forecasting and real-time information systems to improve the sustainability, efficiency and safety of mobility and logistics networks (BMDV, 2018[158]). Through its Sustainable Mobility initiative, which is part of its NRRP, Italy is seeking to optimise environmental and energy resources through AI-driven traffic and transport management (AGID, 2024[166]). France has adopted a broad approach with its Digitisation and Decarbonisation of Mobility strategy, which allocates EUR 570 million to AI-based mobility systems that support sustainability and emissions reduction. The Slovak Republic has embedded AI into its Strategic Transport Development Plan until 2030, aiming to develop a cohesive national transport information system. The plan prioritises AI-driven real-time traffic management; multimodal transport modelling; and automation in rail, road and water transport.
Countries are also supporting AI-driven R&D for mobility transformation. In Austria, the National Funding Call Mobility Transition 2024/1 – Mobility Technologies and Components provides EUR 5 million to projects that align with the country’s sustainability goals. These focus especially on avoiding emissions and shifting to more efficient transport modes (Mobilitätswende, 2024[167]). Czechia has established two key programmes. Transport 2030, a EUR 78 million applied research programme, supports AI in automation, digitalisation and advanced transport technologies (TACR, 2024[168]). Meanwhile, the Mobility Innovation Hub is a national incubator linking start-ups, research institutions and industry to accelerate sustainable mobility solutions (Technology Incubator, 2024[169]). Sweden launched Strategic Vehicle Research and Innovation in 2009, a joint public-private programme investing EUR 86 million per year in sustainable road transport innovations (EC, 2025[170]). The programme finances AI-driven projects in transport electrification, automation and emissions reduction. In Finland, Future Mobility Finland (2025[171]) promotes AI adoption in public and private transport, logistics and sustainable urban mobility projects as part of the country’s National Programme for Sustainable Growth in the Transport Sector.
Although mobility data sharing is essential for AI-driven transport, only three EU Member States have reported targeted initiatives
Despite the relevance of mobility data for AI-driven transport, only three EU Member States have reported initiatives that target the sharing of these data. The ability to share and access mobility data securely is a critical enabler of AI-driven transport systems, supporting real-time traffic management, multimodal mobility services and logistics optimisation. However, only three EU Member States have reported dedicated initiatives to foster mobility data sharing. This indicates that mobility data sharing remains an underdeveloped area in most EU countries, despite its relevance for AI-powered transport solutions.
A key challenge in mobility data sharing is balancing data sovereignty with interoperability, ensuring that private and public sector data can be securely exchanged without centralising sensitive information. Germany has taken the most structured approach with the Mobility Data Space (MDS), a federally funded initiative to facilitate the secure and voluntary exchange of transport data (Mobility Data Space, 2025[172]). Unlike traditional data platforms, the MDS does not store data centrally but allows original holders to control access. The EUR 13.76 million initiative aims to support AI-based mobility services by unlocking access to non-public datasets in a controlled manner.
Another key German initiative, AIAMO Mobility Data Hub and Services, focuses on creating a data-sovereign platform for mobility services. Led by the Federal Ministry of Transport, AIAMO integrates various mobility services, making AI-driven transport solutions more accessible for SMEs and smaller cities. By simplifying data collection and analysis, AIAMO enables insights into multimodal and resilient traffic management, helping smaller mobility operators access AI applications. AIAMO is supported by EUR 22.29 million in funding (BMDV, 2023[173]).
Germany also leads efforts on another priority: open-source data sharing for logistics and supply chains. The Silicon Economy ecosystem project, funded with EUR 33.77 million, provides logistics companies with open-source AI-based data models, interfaces and software components. In so doing, it helps automate business processes and improve supply chain transparency (Silicon Economy, 2025[174]).
Beyond logistics, mobility data integration for urban transport planning is an emerging focus. Ireland is experimenting with AI-based data mapping through Mapping City Assets from Street Level Imagery, part of the Smart Dublin (2020[175]) Initiative. This project collects detailed geospatial data on infrastructure elements such as traffic signs, bins and storm drains, allowing for improved city planning and operations. Italy is working towards a broader national framework with Urban Mobility Data Integration. This initiative will aim to establish a national registry of datasets and models for mobility applications, supporting real-time optimisation of transport networks (AGID, 2024[166]).
The small number of countries investing in mobility data-sharing initiatives suggests that data availability and interoperability are not yet prioritised as much as AI-driven automation or digital mobility services. As AI applications in mobility continue to expand, the lack of structured data-sharing frameworks across most EU Member States could become a bottleneck for future innovation.
A third of EU Member States have launched AI-driven initiatives to help transform urban mobility
One‑third of EU Member States have introduced initiatives to support transformation of urban mobility. These 17 efforts reflect growing momentum across diverse sub-domains such as traffic management, public transport and mobility-as-a-service (MaaS), laying the groundwork for more integrated and sustainable urban transport systems.
AI applications for road maintenance and traffic management account for the largest share of initiatives, with projects in Ireland, Italy, Latvia, Malta, the Netherlands and Slovenia:
Ireland is piloting AI-enabled cameras to analyse vehicular flows in real time, allowing authorities to monitor traffic patterns and improve road safety (Smart Dublin, 2020[175]).
In Italy, AI for Smart Cities Mobility focuses on using AI to optimise traffic flow in high-density areas and improve public transport operations (AGID, 2024[166]).
Latvia has introduced four AI-driven traffic management systems. Winter Road Maintenance Using AI assesses road conditions in real time to optimise maintenance (Labs of Latvia, 2023[176]). The LMT traffic monitoring tool, an AI-powered General Data Protection Regulation-compliant video technology, detects traffic intensity and violations such as red-light running and improper lane use (LMT, 2025[177]). The country has also deployed smart LED lighting, which integrates AI-driven adaptive lighting solutions that monitor road conditions and violations to enhance urban safety. Lastly, the 3visionD (2025[178]) project employs AI-powered video analytics to manage urban parking, monitor public spaces and implement automated traffic control measures.
The AI for Traffic Management Pilot Project in Malta embeds AI into national traffic control systems to reduce congestion and improve journey planning. Meanwhile, Intelligent Transport Systems uses AI-driven route optimisation to improve traffic flow and introduce low-traffic zones (Malta.AI, 2019[179]).
The Netherlands has developed the A12 Traffic Forecaster, a machine-learning application that predicts traffic congestion based on real‑time and historical data, helping commuters make informed route decisions (A12 Slim Reizen, 2024[180]). The country has also launched Smart Traffic Analysis and Traffic Safety, which applies AI and computer vision to improve road safety by monitoring accident-prone intersections and pedestrian crossings (NHL Stenden, 2024[181]).
Slovenia is integrating AI into national traffic management through two initiatives. The National Traffic Management Centre consolidates road and traffic data for improved mobility insights. For its part, the DARS AI Initiative leverages machine learning to predict congestion and accidents on motorways (Slovenian Ministry of Infrastructure, 2022[182]).
Several EU Member States are also using AI to enhance public transport, as well as to promote multimodal mobility and MaaS solutions. In Bulgaria, the INNOAIR project is a demand-responsive transport system using AI to optimise public transport routes dynamically. The initiative includes AI-powered electric minibuses that adjust routes based on passenger demand, improving connectivity for underserved urban areas while reducing emissions. In Portugal, the Cooperative Streets initiative tested Cooperative Intelligent Transport Systems in urban environments, applying AI for multimodal access, parking management and public transport integration (EC, 2025[183]). Through the NetZeroCities (2024[17]) initiative, Slovenia is upgrading its digital transport platform by integrating AI-driven public transport and traffic data. At the same time, it is developing a MaaS mobile application to consolidate urban mobility services.
The Netherlands has launched a unique initiative targeting cyclist safety. BikeSafeAI applies AI‑powered traffic analysis to identify high-risk intersections and road segments for cyclists (SIA, 2024[184]). By mapping dangerous traffic patterns, the system helps city planners improve cycling infrastructure and implement safety measures to prevent accidents.
AI technologies such as deep learning, machine learning and artificial neural networks offer great promise in improving farming practices and enhancing the sustainability of agriculture through data-driven approaches (Rejeb et al., 2022[185]). The ability of AI to process and analyse data from sensors and IoT allows for precise analytics regarding soil fertility, disease diagnostics, irrigation levels and pest control (Lin et al., 2019[186]). Combining computer vision with AI algorithms can result in benefits such as task automation, improved food quality in grains through disease identification and pest infestation detection (Patricio and Rieder, 2018[187]). AI-powered robots and agricultural machines can automate labour-intensive tasks, optimise resource efficiency and increase agricultural output, helping the sector meet its growing demands despite environmental constraints and labour shortages (Rejeb et al., 2022[185]).
As one of the world’s leading agrifood players, the European Union faces a triple challenge of ensuring food security and nutrition; supporting the livelihood of farmers and other stakeholders in the agricultural supply chain; and improving environmental resilience (OECD, 2023[188]). Changing weather conditions exacerbate these challenges, threatening agricultural production and quality, particularly for fruit crops. (EC, 2024[189]).
In response to these challenges, recent years have witnessed an increased adoption of automation tools, including precision farming, automated feeders, drop irrigation systems and mechanised harvesting (EC, 2024[189]). The digitalisation of agriculture in the European Union stands as a combination of wireless technologies, IoT, AI and blockchain under the umbrella of precision farming (Kondratieva, 2021[190]). AI is being combined with traditional agricultural methods and IoT to improve agriculture by monitoring crop growth using the principles of precision farming (Elbasi and Mostafa, 2023[191]). This integration can enable farmers to increase crop yields, control pests, monitor soil conditions, manage workloads, predict optimal time for sowing and harvesting, and improve agricultural tasks in the food supply chain (Zhou and Chen, 2023[192]).
The EU Coordinated Plan on AI has identified AI and digital technologies as a key enabler for agriculture, emphasising its potential to increase farm efficiency and economic and environmental sustainability. 25 EU Member States signed a declaration in 2019 to foster a smart and sustainable digital future for European agriculture and rural areas. Since 2020, the European Commission has allocated EUR 175 million to Horizon 2020 research projects to digitalise agriculture, focusing on deployment of digital technologies such as AI, robotics and IoT.
The EU Coordinated Plan on AI outlines several key actions for Member States to help leverage AI in agriculture:
Take full advantage of RRF funding for the digitalisation of the agri-food sector, as envisaged in the national plans, for example to set up additional AI and robotics TEFs and EDIHs in agri-food, in addition to those already planned under the Digital Europe programme.
Take an active role in the partnership Agriculture of Data.
Consider funding of national R&I projects that link AI and robotics technologies to their use in agriculture, forestry, rural development and bioeconomy.
Initiatives to foster adoption of AI in agriculture are expanding among EU Member States
Two-thirds of EU Member States – 18 of 27 countries – have launched initiatives to foster adoption of AI in agriculture (Figure 5.8). While this represents significant progress, the distribution of these efforts remains uneven across countries. This disparity arises from factors such as geographical features, barriers to widespread AI adoption, farm type and size, agricultural output and levels of digital literacy. Additionally, the focus on agriculture varies depending on the sector’s economic importance within each country. Countries where agriculture plays a central role in the economy are more likely to prioritise AI integration into farming practices.
Initiatives focused on AI in agriculture within the European Union primarily emphasise fostering scientific and industrial research. In all, 25 initiatives across 14 EU Member States advance AI-driven technologies for agriculture at a foundational level. Additionally, seven EU Member States report 11 initiatives to transfer these technologies and knowledge to farmers. While fewer in number, initiatives promoting agricultural data sharing are also gaining momentum, highlighting growing recognition of how data are essential in leveraging AI for agricultural transformation. Lastly, initiatives focusing on livestock farming and crop management practices are limited to three Member States only.
Fourteen EU Member States have reported 25 initiatives to foster scientific and industrial research in agrifood technologies. Reported initiatives focus on testing and innovation hubs, institution-led projects and ministry-led cross-sectoral collaboration. In Austria, for example, the AI for Green initiative promotes R&D projects that develop new AI technologies and contribute to the country’s climate goals (FFG, 2025[193]). In Belgium, the Flanders Research Institute for Agriculture, Fisheries and Food (ILVO) focuses on AI-driven precision farming (Vangeyte, 2025[194]) and agricultural robotics use-cases through its Flanders AI Research Programme (Flanders AI Research Program, 2025[195]). With a EUR 2 million grant, the Flemish government and ILVO co-fund the testing and experimentation facility on AI in agrifood. Over 2023‑2027, Bulgaria aims to support 120 operational groups on innovative projects related to use of AI as part of its Strategic Plan for the Development of Agriculture and Rural Areas (Government of Bulgaria, 2023[196]).
Source: Data reported by EU Member States through the survey and interviews.
With a budget of EUR 32.4 million, the Ministry of Agriculture in Czechia supports the purchase of AI-powered agricultural robotics as part of its Research, Development and Innovation Strategy (2023‑2032) (Ministry of Agriculture of the Czech Republic, 2022[197]). Denmark funds various R&D projects through its Grønt Udviklings- og Demonstrationsprojekt programme (GUDP, 2025[198]).
EU Member States have widely funded scientific and industrial research to develop AI-driven agriculture technologies, with 14 countries having launched initiatives
Fourteen EU Member States have launched research initiatives to develop AI-driven agriculture technologies. In Hungary, as part of the national AI Strategy, the Ministry of Agriculture supports projects for predictive analytics and optimised plant and livestock management (EC, 2021[199]). In Ireland, Crop Optimisation through Sensing, Understanding and Visualisation (CONSUS) is a EUR 17.6 million five-year strategic research programme focused on digital agriculture. The research of several other EU Member States is highlighted below.
Along with Austria and France (which participates with EUR 30 million funding), Germany is part of the European Testing and Experimentation Facilities for Agrifood Innovation. This fosters a network of digital facilities and helps assess and validate AI and robotics solutions in real-world conditions (agrifoodTEF, 2025[200]). As part of its national AI strategy, Germany initiated 36 agriculture research projects in February 2020 (EUR 44 million) (BMEL, 2025[201]).
The Netherlands promotes use of AI in the agrifood sector through several government initiatives:
The Knowledge and Innovation Agenda for Agriculture, Water and Food (KIA LWV) supports development of technologies such as sensors, robots and digital twins through three dedicated innovation programmes (KIA LWV, 2024[202]).
The AI-hub Noord-Nederland focuses on sharing data, fostering knowledge about digitalisation and AI, and developing autonomous systems (NL AIC, 2024[203]).
The NXTGEN Hightech (2025[204]) programme develops autonomous agricultural machinery.
The Action Program Digitalisation of the Ministry of Agriculture, Fisheries, Food Security and Nature (LVVN) supports AI systems addressing smart farming and food processing through its EUR 52.7 million (2023‑2029) (LVVN, 2023[205]).
The Farm of the Future initiative in Lelystad – a pilot programme of the European Innovation Partnership FieldLabs – demonstrates circular agriculture solutions with AI applications through collaboration between Wageningen University & Research (WUR) and Dutch farmers (Boerderij van de Toekomst, 2025[206]). From 2019 to 2024, the LVVN funded the WUR (2025[207]) on various research on AI in animal and arable systems.
In Portugal, the national AI Strategy supports R&D projects focused on AI solutions for agriculture through the COMPETE 2020 (2025[208]) programme (EUR 12.9 million). Fraunhofer Portugal AWAM, the country’s R&D centre for agriculture, has been researching the potential role of AI in agriculture since 2018 (Portugal INCoDe.2030, 2022[209]; Fraunhofer Portugal, 2025[210]).
Romania invests in its Research Institute for Artificial Intelligence as an integral part of the Romanian AI Hub (Lucuț, 2022[211]), supporting digital and robotic technologies in agriculture. Its national AI strategy supports research into automation solutions, AI-driven robotics and R&D projects under the Horizon Europe programme. The ADER 2026 programme (EUR 39.2 million) funds research on climate adaptation, biodiversity conservation and implementation of innovative technologies (Secretariatul General al Guvernului, 2024[212]).
The Slovak Republic, through the ICT-AGRI-FOOD2024 Joint Call and the Smart Villages initiative, supports projects that transform agrifood systems using ICT technologies and AI for precision farming, smart resource management and enhancing rural services such as healthcare and education.
In Spain, the EUR 10 million AgrarIA consortium project develops AI solutions for intelligent agriculture (GMV, 2022[213]).
Some EU Member States are recognising the importance of knowledge and technology transfer of AI solutions in agriculture
Seven EU Member States have reported 11 initiatives to support the knowledge and technology transfer of AI solutions, mainly through the EDIH network and similar knowledge hubs (Table 5.8).
Source: Data reported by EU Member States through the survey and interviews.
AI is being integrated to boost efficiency in livestock farming, crop and aerial management
Several countries are integrating AI to boost efficiency in farming. In Germany, the Development of a Remote Sensing and Model-based Decision Support System for Sustainable Agriculture project aims to aid farmers in sustainable decision making. It uses an integrated system that combines remote sensing and modelling (FNR, 2025[214]). In Ireland, the VistaMilk (2025[215]) SFI Research Ireland Centre uses AI to improve sustainability in dairy farming practices. The Slovak Republic invested EUR 3.5 million to build a fully operational aerial monitoring system. It monitors agricultural land with the potential to expand to agricultural production when integrated with commercial AI solutions.
Agriculture data sharing is essential for better AI-driven agricultural practices, yet only five EU Member States have reported targeted initiatives
While sharing of agriculture data is essential to improve AI-driven agriculture, only five EU Member States have reported initiatives in this area. In Belgium, ILVO invests in the data-sharing project DjustConnect (2025[216]) that returns control to the data producer. The Flemish government also co-funds participation in the Agricultural Data Space at the EU level. France has financed the Agdatahub, Intermédiaire de données du secteur agri et agro, a platform for agricultural and agrifood data. It unites public and private players to facilitate the digital transformation of agriculture based on real-world use-cases. In Hungary, the Agri-Data Framework enhances structuring and accessibility of agricultural data. In Portugal, AGRISPACE – part of the national consortium Portuguese AGRIculture Data SPACE – is developing a national data space for agriculture (Diário da República, 2024[217]). Slovenia supports agricultural data spaces for AI through two projects: Green.Dat.AI (EUR 6.66 million) for smart farming optimisation through digital twins and water management techniques. DS2 – DataSpace, DataShare 2.0 – a modular software facilitates compliance of precision agriculture data sharing with EU regulations.
References
[178] 3visionD (2025), “Smart parking for smart city”, https://3visiond.com/parking/ (accessed on 4 May 2025).
[180] A12 Slim Reizen (2024), “Homepage”, https://www.a12slimreizen.nl/ (accessed on 6 November 2024).
[77] Advantis Medical Imaging (2024), “Explore Advantis platform”, https://advantis.io/ (accessed on 14 November 2024).
[166] AGID (2024), “The Italian strategy for artificial intelligence 2024-2026”, Agenzia per l’Italia Digitale, https://www.agid.gov.it/en/agenzia/stampa-e-comunicazione/notizie/2024/07/22/italian-strategy-artificial-intelligence-2024-2026 (accessed on 27 February 2025).
[200] agrifoodTEF (2025), “The European testing and experimentation facilities for agrifood innovation”, https://www.agrifoodtef.eu/ (accessed on 2 December 2024).
[31] AI Hub Plastic Packaging (2024), “AI application hub on plastic packaging”, https://ki-hub-kunststoffverpackungen.de/en/ (accessed on 15 November 2024).
[79] AI Sweden (2024), “A shared digital assistant for the public sector”, https://www.ai.se/en/project/shared-digital-assistant-public-sector (accessed on 27 January 2025).
[70] AI4Health.Cro (2021), “Networking services”, European Digital Innovation Hub, https://ai4healthcro.eu/?ht_kb=networking-services (accessed on 8 January 2025).
[73] AIDA (2024), “AIDA data hub definition”, https://datahub.aida.scilifelab.se/policy/definition/ (accessed on 3 March 2025).
[14] AiNed (2024), “Homepage”, https://ained.nl/en (accessed on 6 November 2024).
[99] AK Wien (2024), “Xplain-AI”, Arbeiterkammer Wien, https://wien.arbeiterkammer.at/service/digifonds/gefoerderte-projekte/Xplain_AI.html (accessed on 18 March 2025).
[151] ALP.Lab (2024), “About us”, https://www.alp-lab.at/about-us/ (accessed on 8 April 2025).
[15] AMA (2023), Estratégia Nacional de Territórios Inteligentes, [English translation], Área Governativa da Digitalização e da Modernização Administrativa, https://www.ama.gov.pt/documents/24077/320185/ENTI_RCM_V2_atualizado.pdf/634a2167-0e64-48e5-9ddc-ef1fd3fc0ca8 (accessed on 12 April 2025).
[91] AmCham Slovakia (2024), “Common recommendations for the healthcare system 2030”, https://amcham.sk/publications/issues/2020-2-innovations-in-the-digital-age/article/273287/common-recommendations-for-the-healthcare-system-2030 (accessed on 7 January 2025).
[9] Barnabas, A. and E. Owen (2025), “How AI is shaping the future of carbon accounting in ESG reporting”, Preprints, https://doi.org/10.20944/preprints202503.0515.v1 (accessed on 25 April 2025).
[148] Bathla, G. et al. (2022), “Autonomous vehicles and intelligent automation: Applications, challenges, and opportunities”, Mobile Information Systems, Vol. 1, p. 7632892, https://doi.org/10.1155/2022/7632892.
[13] BBK (2024), “Athens, Thessaloniki, Kalamata, Ioannina, Trikala and Kozani in “100 Climate Neutral Cities until 2030””, Construction Development, https://bbk.gr/en/athens-thessaloniki-kalamata-ioannina-trikala-and-kozani-in-100-climate-neutral-cities-until-2030/ (accessed on 1 February 2025).
[48] BKA (2024), “Änderung des Allgemeinen Sozialversicherungsgesetzes (BGBl. I Nr. 190/2023)”, [English translation], Legal Information System of the Republic of Austria, https://ris.bka.gv.at/eli/bgbl/I/2023/190 (accessed on 16 February 2025).
[90] BKUS (2024), “Automated voice communication solutions for the health sector”, Bērnu klīniskā universitātes slimnīca, https://www.bkus.lv/lv/automatizeti-balss-komunikacijas-risinajumi-veselibas-nozarei (accessed on 3 February 2025).
[173] BMDV (2023), “Artificial intelligence and mobility – AIAMO”, Bundesministerium für Digitales und Verkehr, https://bmdv.bund.de/SharedDocs/DE/Artikel/DG/KI-Projekte/aiamo.html (accessed on 4 November 2024).
[159] BMDV (2021), “Research funding for ensuring a viable and sustainable mobility system through automation and connectivity”, German Federal Ministry of Transport, https://bmdv.bund.de/SharedDocs/EN/Articles/DG/research-programme-on-automation-and-connectivity-in-road-transport.html (accessed on 13 August 2024).
[158] BMDV (2018), KI-Projekte des BMDV, [English translation], German Federal Ministry of Transport, https://bmdv.bund.de/DE/Themen/Digitales/Kuenstliche-Intelligenz/KI-Projekte-in-der-Mobilitaet/aktionsplan.html (accessed on 17 March 2025).
[201] BMEL (2025), “BMEL fördert Projekte zum Einsatz von Künstlicher Intelligenz in der Landwirtschaft und den ländlichen Räumen”, [English translation], Federal Ministry of Food and Agriculture, https://www.bmel.de/DE/themen/digitalisierung/kuenstliche-intelligenz.html (accessed on 9 April 2025).
[120] BMG (2024), “genomeDE – national strategy for genomic medicine”, German Federal Ministry of Health, https://www.bundesgesundheitsministerium.de/en/en/international/european-health-policy/genomde-en.html (accessed on 7 February 2025).
[50] BMG (2024), “Gesundheitsdatennutzungsgesetz (GDNG)”, [English translation]. Bundesministerium für Gesundheit, https://www.bundesgesundheitsministerium.de/service/gesetze-und-verordnungen/detail/gesundheitsdatennutzungsgesetz.html (accessed on 7 May 2025).
[136] BMI (2024), “Künstliche Intelligenz in der Verwaltung”, [English translation], Federal Ministry of the Interior of Germany, https://www.cio.bund.de/Webs/CIO/DE/digitale-loesungen/datenpolitik/daten-und-ki/daten-und-ki-node.html (accessed on 4 May 2025).
[10] BMUV (2024), “Künstliche Intelligenz für Umwelt und Klima”, [English translation], Federal Ministry for the Environment and Climate Action, https://www.bmuv.de/themen/digitalisierung/kuenstliche-intelligenz/kuenstliche-intelligenz-fuer-umwelt-und-klima (accessed on 6 May 2025).
[27] BMUV (2024), “Unsere Initiative “Green-AI Hub Mittelstand””, [English translation], German Federal Ministry for the Environment and Consumer Protection, https://www.bmuv.de/themen/digitalisierung/kuenstliche-intelligenz/kuenstliche-intelligenz-fuer-umwelt-und-klima/unsere-initiative-green-ai-hub-mittelstand (accessed on 6 May 2025).
[157] BMWK (2023), “The federal government’s lightweighting strategy”, German Federal Ministry for Economic Affairs and Climate Action, https://www.bmwk.de/Redaktion/EN/Publikationen/Schluesseltechnologien/the-federal-governments-lightweighting-strategy.pdf?__blob=publicationFile&v=3 (accessed on 26 October 2024).
[206] Boerderij van de Toekomst (2025), “Farm of the future”, https://farmofthefuture.nl/ (accessed on 13 May 2025).
[76] CellaVision (2024), “CellaVision in a nutshell”, https://www.cellavision.com/ (accessed on 14 November 2024).
[133] CGA (2023), “CGA”, Caixa Geral de Aposentações, https://www.cga.pt/cgainicio.asp (accessed on 1 March 2025).
[135] Chamber of Deputies (2024), “La Chambre se dote d’une Charte sur l’intelligence artificielle”, Grand-Duché de Luxembourg, https://www.chd.lu/fr/charteIA (accessed on 6 November 2024).
[6] Chen, L. et al. (2023), “Artificial intelligence-based solutions for climate change: A review”, Environmental Chemistry Letters, Vol. 21/5, pp. 2525-2557, https://doi.org/10.1007/s10311-023-01617-y.
[208] COMPETE 2020 (2025), “Homepage”, https://www.compete2020.gov.pt/ (accessed on 2 December 2024).
[80] Corti (2024), “The trusted LLM for higher quality care”, https://www.corti.ai/ (accessed on 14 November 2024).
[149] Craglia, M. et al. (2021), “AI Watch, AI Uptake in Smart Mobility”, https://doi.org/10.2760/879190 (accessed on 21 October 2024).
[128] CY-Biobank (2024), “The center”, https://biobank.cy/profile/ (accessed on 6 November 2024).
[156] Czech Ministry of Transportation (2024), “Ethics commission”, https://md.gov.cz/Uzitecne-odkazy/Autonomni-mobilita/Eticka-komise-Link?returl=/Uzitecne-odkazy/Autonomni-mobilita (accessed on 16 January 2025).
[64] Datos.gob.es (2024), “National health data space: A strategic project for the country”, https://datos.gob.es/en/blog/national-health-data-space-strategic-project-country (accessed on 3 April 2025).
[145] de Queiroz, A. et al. (2021), “Overview of urban mobility in smart cities”, Research, Society and Development, Vol. 10/9, pp. e18210917830-e18210917830, https://doi.org/10.33448/rsd-v10i9.17830.
[51] Deutscher Ärzteverlag (2023), “Registergesetz: Eckpunkte stoßen auf Zustimmung”, Deutsches Ärzteblatt, [English translation], Deutsches Ärzteblatt GmbH, https://www.aerzteblatt.de/archiv/232457/Registergesetz-Eckpunkte-stossen-auf-Zustimmung (accessed on 9 February 2025).
[24] Deutscher Wetterdienst (2025), “CONTRAILS project”, https://www.dwd.de/DE/forschung/projekte/contrails/contrails_node.html (accessed on 26 March 2025).
[217] Diário da República (2024), “Despacho n.º 3073/2024, de 22 de março”, [English translation], https://diariodarepublica.pt/dr/detalhe/despacho/3073-2024-856836196 (accessed on 9 May 2025).
[137] Digital Wallonia (2022), “Stratégie Digital Wallonia. Ambition “Administration Numérique””, https://www.digitalwallonia.be/fr/publications/ambition-administration-numerique/ (accessed on 18 October 2024).
[41] Digitaliseringsstyrelsen (2025), “Green data processing and storage”, https://digst.dk/digital-transformation/digital-groen-omstilling/groen-databehandling-og-opbevaring/ (accessed on 7 February 2025).
[152] Digitrans (2024), “Best answers for future transport”, https://www.digitrans.expert/en/ (accessed on 5 November 2024).
[104] Digizo.nu (2024), “Our shoulders to the wheel of healthcare transformation”, https://digizo.nu/ (accessed on 8 November 2024).
[216] DjustConnect (2025), “DjustConnect”, https://www.djustconnect.be/en (accessed on 10 January 2025).
[155] Drive Sweden (2024), “Projects”, https://www.drivesweden.net/en/projects (accessed on 14 April 2025).
[38] DS2 (2023), “Trusted data exchange across sectors”, https://www.dataspace2.eu/ (accessed on 8 January 2025).
[110] DS4Health (2024), “Homepage”, Digital Skill for Healthcare Transformation, https://www.digitalskills4health.eu/kopie-von-digital-health-hub (accessed on 8 November 2024).
[141] Du Plessis, M. et al. (2025), “Shaping the future of freight logistics: Use cases of artificial intelligence”, Sustainability, Vol. 17/4, p. 1355, https://doi.org/10.3390/su17041355.
[55] Dutch Ministry of Health, Welfare and Sport (2024), “Electronic data exchange in healthcare act (Wegiz)”, https://www.datavoorgezondheid.nl/wegiz (accessed on 3 May 2025).
[219] Dutch Ministry of Health, Welfare and Sport (2024), EU Interoperability with HERA’s IT Platform, https://www.rivm.nl/en/international-projects/eu-hip (accessed on 3 May 2025).
[45] Dutch Ministry of Health, Welfare and Sport (2024), “Health data access body (HDAB)”, https://www.datavoorgezondheid.nl/health-data-acces-body (accessed on 3 May 2025).
[183] EC (2025), “Cooperative streets”, European Commission, https://trimis.ec.europa.eu/project/cooperative-streets (accessed on 3 April 2025).
[170] EC (2025), “Sweden’s collective expertise within innovation available for companies and public administrations”, European Commission, https://european-digital-innovation-hubs.ec.europa.eu/edih-catalogue/countries/sweden (accessed on 2 May 2025).
[129] EC (2024), “Bulgaria’s recovery and resilience plan”, European Commission, https://data.europa.eu/doi/10.2792/965801 (accessed on 15 March 2025).
[53] EC (2024), “Enabling patient access to their health data: PATHED”, European Commission, https://www.developmentaid.org/organizations/awards/view/440805/enabling-patient-access-to-their-health-data-pathed (accessed on 28 January 2025).
[189] EC (2024), EU Agricultural Outlook 2023-2035, European Commission, Brussels, https://agriculture.ec.europa.eu/system/files/2024-01/agricultural-outlook-2023-report_en_0.pdf.
[103] EC (2022), “Establishment of a health professionals competence platform”, European Commission, https://commission.europa.eu/projects/establishment-health-professionals-competence-platform_en (accessed on 22 September 2024).
[199] EC (2021), “Hungary AI strategy report”, European Commission, https://ai-watch.ec.europa.eu/countries/hungary/hungary-ai-strategy-report_en (accessed on 23 October 2024).
[82] EC (2021), “Slovakia’s recovery and resilience plan”, European Commission, https://commission.europa.eu/business-economy-euro/economic-recovery/recovery-and-resilience-facility/country-pages/slovakias-recovery-and-resilience-plan_en?locale=en (accessed on 7 January 2025).
[42] Ecolab (2024), Référentiel général pour l’IA frugale. Une AFNOR Spec pour mesurer et réduire l’impact environnemental de l’IA, Ecolab, Paris, https://greentechinnovation.fr/storage/2024/06/Referentiel-general-pour-lIA-frugale.pdf.
[54] E-hälsomyndigheten (2024), “Förslag till färdplan för genomförandet av en nationell digital infrastruktur för hälsooch sjukvårde”, [English translation], E-hälsomyndigheten, https://www.ehalsomyndigheten.se/globalassets/ehm/3_om-oss/rapporter/rapporter_regeringsuppdrag/fardplan/forslag-till-fardplan-for-genomforandet-av-en-nationell-digital-infrastruktur-for-halso--och-sjukvarden---slutredovisning.pdf.
[62] EIBIR (2024), “Building a data-driven future for cancer care”, European Institute for Biomedical Imaging Research, https://cancerimage.eu/ (accessed on 13 November 2024).
[57] EIT Health (2024), “European health data space in Portugal”, White Paper, EIT Health, Coimbra, https://eithealth.eu/wp-content/uploads/2024/06/EHDS-white-paper-Portugal.pdf.
[191] Elbasi, E. and N. Mostafa (2023), “Artificial intelligence technology in the agricultural sector: A systematic review”, IEEE Access, Vol. 11/1, pp. 171-202, https://doi.org/10.1109/access.2022.3232485.
[125] ELIXIR (2024), “Homepage”, https://elixir-europe.org/ (accessed on 6 November 2024).
[56] Esveikata.lt (2023), “LNKC service development project”, https://www.esveikata.lt/lnkc-paslaugu-vystymo-projektas (accessed on 20 October 2024).
[49] EU (2025), “Regulation (EU) 2025/327 of the European Parliament and of the Council of 11 February 2025 on the European Health Data Space and amending Directive 2011/24/EU and Regulation (EU) 2024/2847”, European Parliament and Council of the European Union, https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ%3AL_202500327.
[105] EU (2021), “Digivaardig in de Zorg”, [English translation], European Commission, https://digital-skills-jobs.europa.eu/en/inspiration/good-practices/digivaardig-de-zorg (accessed on 22 February 2025).
[75] EuCanImage (2024), “Homepage”, https://eucanimage.eu/ (accessed on 14 November 2024).
[19] Europol (2024), “SIRIUS project”, https://www.europol.europa.eu/operations-services-innovation/sirius-project (accessed on 30 April 2025).
[144] Fernández-Llorca, D. and E. Gómez (2023), “Trustworthy artificial intelligence requirements in the autonomous driving domain”, Computer, Vol. 56/2, pp. 29-39, https://doi.org/10.1109/MC.2022.3212091.
[193] FFG (2025), “AI Ökosysteme”, https://www.ffg.at/ai (accessed on 15 November 2024).
[88] FFG (2024), “We connect Austria’s care and assistance”, https://www.linkedcare.at/en/ (accessed on 14 November 2024).
[161] FFI (2025), “FFI safe automated driving”, https://ffisweden.se/en/ffi-traffic-safe-automation/ (accessed on 3 May 2025).
[66] Findata (2025), “FinHITS – strengthening Finnish health data ICT for secondary use”, https://findata.fi/en/finhits/ (accessed on 29 January 2025).
[46] Findata (2024), “Homepage”, Finnish Social and Health Data Permit Authority, https://findata.fi/en/ (accessed on 13 November 2024).
[39] Finnish Ministry of Transport and Communications (2021), “Finland announces climate strategy for ICT sector: Harnessing data bits to combat climate change”, https://lvm.fi/en/-/finland-announces-climate-strategy-for-ict-sector-harnessing-data-bits-to-combat-climate-change-1260885 (accessed on 18 August 2024).
[195] Flanders AI Research Program (2025), “Robotics for agrifood”, https://www.flandersairesearch.be/en/research/use-cases/robotics-for-agrifood (accessed on 23 March 2025).
[20] Flanders AI Research Program (2024), “Improved energy production of wind turbine parks”, https://www.flandersairesearch.be/en/research/use-cases/improved-energy-production-of-wind-turbine-parks (accessed on 23 March 2025).
[214] FNR (2025), “Verbundvorhaben: Optimierung von Datenerfassung und Steuerungstechnik für Biomassefeuerungen; Teilvorhaben 1: Theoretische und experimentelle Untersuchungen, Entwicklung Module zur Datenerfassung - Akronym: DigitalFire”, [English translation], Fachagentur Nachwachsende Rohstoffe e.V., https://www.fnr.de/index.php?id=11150&fkz=22039618 (accessed on 4 January 2025).
[65] Foundation 29 (2024), “Homepage”, https://www.healthdata29.org/ (accessed on 13 November 2024).
[84] FrailCare.AI (2024), “Homepage”, Fundação para a Ciência e a Tecnologia, https://frailcareai.vohcolab.org/ (accessed on 14 November 2024).
[210] Fraunhofer Portugal (2025), “About Fraunhofer Portugal”, https://www.awam.fraunhofer.pt/en/about-us.html (accessed on 10 March 2025).
[95] French Government (2021), Stratégie d’accélération « Santé numérique », French Ministry of Solidarités and Health, https://www.economie.gouv.fr/files/files/2021/DP_sante_numerique_20211019.pdf.
[171] Future Mobility Finland (2025), “Homepage”, https://futuremobilityfinland.fi/ (accessed on 10 January 2025).
[142] Garikapati, D. and S. Shetiya (2024), “Autonomous vehicles: Evolution of artificial intelligence and the current industry landscape”, Big Data and Cognitive Computing, Vol. 8/4, p. 42, https://doi.org/10.3390/bdcc8040042.
[117] GDHP (2024), “Who’s involved”, Global Digital Health Partnership, https://gdhp.health/about/whos-involved/ (accessed on 20 March 2025).
[115] GDI (2024), “Who is involved”, European Genomic Data Infrastructure, https://gdi.onemilliongenomes.eu/about/participants (accessed on 6 November 2024).
[60] gematik GmbH (2024), “National interoperability governance”, https://www.ina.gematik.de/en/interoperability-governance (accessed on 14 February 2025).
[1] George, A., A. George and A. Martin (2024), “The environmental impact of AI: A case study of water consumption by Chat GPT”, Partners Universal International Innovation Journal, Vol. 1/2, pp. 97-104, https://doi.org/10.5281/zenodo.7855594.
[126] GMS (2024), “About us”, Genomic Medicine Sweden, https://genomicmedicine.se/en/about-us/ (accessed on 15 March 2025).
[213] GMV (2022), “Un consorcio liderado por GMV impulsa la investigación de la aplicación de la IA en la cadena de producción agroalimentaria”, [English translation], https://www.gmv.com/es-es/comunicacion/noticias/consorcio-liderado-gmv-impulsa-investigacion-aplicacion-ia-cadena-produccion (accessed on 2 November 2024).
[134] GOV.pl (2025), “mObytawel”, https://www.gov.pl/web/mobywatel (accessed on 13 January 2025).
[196] Government of Bulgaria (2023), “Стратегически план за развитие на земеделието и селските райони за периода 2023-2027 г.”, [English translation], https://sp2023.bg/index.php/bg/sprzsr-bg/za-sprzsr (accessed on 26 February 2025).
[124] Government of Ireland (2021), “Ireland’s national recovery and resilience plan 2021”, https://www.gov.ie/en/publication/d4939-national-recovery-and-resilience-plan-2021/ (accessed on 19 November 2024).
[138] Government of Luxembourg (2024), “GovTech Lab”, https://govtechlab.public.lu/en.html (accessed on 6 November 2024).
[86] Government of Malta (2024), “Pharmacy of your choice”, https://health.gov.mt/poyc/ (accessed on 7 January 2025).
[83] Government of Malta (2023), “The pharmacy of your choice national scheme”, https://poyc.gov.mt/en/poyc-scheme/the-pharmacy-of-your-choice-national-scheme/ (accessed on 7 January 2025).
[23] Government of Poland (2025), “Nowa Energia”, https://www.gov.pl/web/nfosigw/nowa-energia (accessed on 9 March 2025).
[97] Government of Romania (2023), “Health program”, https://oportunitati-ue.gov.ro/en/program/programul-sanatate/ (accessed on 7 April 2025).
[96] Government of Spain (2021), Digital Health Strategy: National Health System, General Secretariat for Digital Health, Government of Spain, https://www.sanidad.gob.es/ciudadanos/pdf/Digital_Health_Strategy.pdf.
[198] GUDP (2025), “GUDP”, Danish Ministry for Food, Agriculture and Fisheries, https://gudp.dk/ (accessed on 7 April 2025).
[67] Health Data Hub (2024), “Homepage”, https://www.health-data-hub.fr/ (accessed on 13 November 2024).
[109] HelloAI (2024), “HelloAI program – AI for healthcare”, https://www.helloaiprofessional.com/ (accessed on 8 November 2024).
[68] ICT & Health (2024), “Belgian universities build a digital platform for Alzheimer research”, https://ictandhealth.com/news/belgian-universities-build-a-digital-platform-for-alzheimer-research (accessed on 9 May 2025).
[59] iED (2024), “Health hub”, Institute of Entrepreneurship Development, https://ied.eu/about/health-hub/ (accessed on 9 May 2025).
[114] IMDRF (2024), “Artificial intelligence/machine learning-enabled”, International Medical Device Regulators Forum, https://www.imdrf.org/working-groups/artificial-intelligencemachine-learning-enabled (accessed on 9 May 2025).
[35] InnoDays (2024), “Climate change cockpit challenge”, https://www.innodays.org/challenge/climate-cockpit/ (accessed on 5 November 2024).
[33] Inria (2025), “OcéanIA”, Institut national de recherche en informatique et en automatique, https://www.inria.fr/fr/oceania (accessed on 10 May 2025).
[165] Instituto de Telecomunicações (2025), “Project: Route 25: Agenda for autonomous, intelligent, interoperable and inclusive mobility”, https://www.it.pt/Projects/Index/4824 (accessed on 29 April 2025).
[7] Kar, A., S. Choudhary and V. Singh (2022), “How can artificial intelligence impact sustainability: A systematic literature review”, Journal of Cleaner Production, Vol. 376, p. 134120, https://doi.org/10.1016/j.jclepro.2022.134120.
[202] KIA LWV (2024), “Kennis- en Innovatieagenda Landbouw-Water-Voedsel 2024-2027”, [English translation], KIA Landbouw, Water en Voedsel, https://www.kia-landbouwwatervoedsel.nl/wp-content/uploads/2024/03/KIA-Landbouw-Water-Voedsel-2024-2027-incl-bijlagen.pdf.
[12] KI-Ideenwerkstatt (2025), KI-Ideenwerkstatt für Umweltschutz, [English translation], https://www.ki-ideenwerkstatt.de/ (accessed on 4 March 2025).
[190] Kondratieva, N. (2021), “EU agricultural digitalization decalogue”, Herald of the Russian Academy of Sciences, Vol. 91/6, pp. 736-742, https://doi.org/10.1134/S1019331621060150.
[108] Kunskapsguiden (2024), “E-hälsa”, [English translation], https://kunskapsguiden.se/omraden-och-teman/verksamhetsutveckling/e-halsa/ (accessed on 10 January 2025).
[176] Labs of Latvia (2023), “Kuldīgas novadā pašvaldības ceļu tīrīšanu un kaisīšanu uzraudzīs mākslīgais intelekts”, [English translation], https://labsoflatvia.com/aktuali/celu-tirisanu-uzraudzis-maksligais-intelekts (accessed on 7 November 2024).
[32] Labs of Latvia (2023), “WinGo Deposit develops state-of-the-art e-cigarette recycling facility”, https://labsoflatvia.com/en/news/wingo-deposit-develops-state-of-the-art-e-cigarette-recycling-facility (accessed on 7 November 2024).
[72] Lantzsch, H. et al. (2022), “Digital health applications and the fast-track pathway to public health coverage in Germany: Challenges and opportunities based on first results”, BMC Health Services Research, Vol. 22/1, p. 1182, https://doi.org/10.1186/s12913-022-08500-6.
[122] Latvian Biomedical Research and Study Center (2020), “Development of genome reference of Latvian population”, https://biomed.lu.lv/project/4-1-1-r-0-3-22-i-vm-001-2/ (accessed on 17 January 2025).
[5] Leal Filho, W. et al. (2022), “Deploying artificial intelligence for climate change adaptation”, Technological Forecasting and Social Change, Vol. 180, p. 121662, https://doi.org/10.1016/j.techfore.2022.121662.
[3] Ligozat, A. et al. (2021), “Unraveling the hidden environmental impacts of AI solutions for environment life cycle assessment of AI solutions”, Sustainability, Vol. 14/9, p. 5172, https://doi.org/10.48550/arXiv.2110.11822.
[186] Lin, Y. et al. (2019), “SensorTalk: An IoT device failure detection and calibration mechanism for smart farming”, Sensors, Vol. 19/21, p. 4788, https://doi.org/10.3990/s19214788.
[177] LMT (2025), “A single traffic monitoring tool for efficient results”, https://trafficmonitoringtech.lmt.lv/ (accessed on 3 March 2025).
[211] Lucuț, C. (2022), “Centrul de Inteligență Artificială al UTCN, aprobat de Guvern! Clădirea va fi construită în Observator”, [English translation], https://www.monitorulcj.ro/administratie/104092-centrul-de-inteligenta-artificiala-al-utcn-aprobat-de-guvern-cladirea-va-fi-construita-in-observator-foto (accessed on 25 February 2025).
[205] LVVN (2023), “Actieprogramma Digitalisering”, [English translation], https://open.overheid.nl/documenten/1d4dc0e6-d437-41f0-a525-709afad726ad/file (accessed on 18 March 2025).
[179] Malta.AI (2019), Malta – The Ultimate AI Launchpad: A Strategy and Vision for Artificial Intelligence in Malta 2030, Malta Digital Innovation Authority, https://malta.ai/wp-content/uploads/2019/11/Malta_The_Ultimate_AI_Launchpad_vFinal.pdf.
[147] Ma, Y. et al. (2020), “Artificial intelligence applications in the development of autonomous vehicles: A survey”, IEEE/CAA Journal of Automatica Sinica, Vol. 7/2, pp. 315-329, https://doi.org/10.1109/JAS.2020.1003021.
[121] Minister of Health of the Republic of Lithuania (2023), Dėl Žmogaus genomo sekoskaitos tyrimų atlikimo tvarkos aprašo patvirtinimo, [English translation], Minister of Health of the Republic of Lithuania.
[197] Ministry of Agriculture of the Czech Republic (2022), “Research, development and innovation strategy of the Ministry of Agriculture for 2023-2032”, https://mze.gov.cz/public/portal/-a30164---YZ5Q7CJN/anglicka-verze-koncepce-vyzkumu-vyvoje-a-inovaci-ministerstva-zemedelstvi-na-obdobi-2023?_linka=a531926 (accessed on 9 April 2025).
[71] Ministry of Foreign Affairs of Denmark (2022), “National Danish research health data gateway assists researchers with accessing data”, https://investindk.com/insights/new-national-danish-research-health-data-gateway-assists-researchers-with-accessing-data (accessed on 8 April 2025).
[98] Ministry of Health of the Republic of Bulgaria (2024), “План за реализация на Националната стратегия за електронно здравеопазване и дигитализация на здравната система 2030 г.”, [English translation], Ministry of Health of the Republic of Bulgaria, https://www.mh.government.bg/upload/12353/Plan%20eZdrave.pdf.
[153] MobiliData (2024), “Remote driving: Test week spring 2024”, https://www.mobilidata.be/en/news/remote-driving-test-week-spring-2024 (accessed on 17 May 2025).
[167] Mobilitätswende (2024), “Mobilitätswende 2024/1 – Mobilitätstechnologie”, [English translation], https://fti-mobilitaetswende.at/de/artikel/ausschreibungen/Mobilitaetswende-2024-1-mobilitaetstechnologie.php (accessed on 27 February 2025).
[172] Mobility Data Space (2025), “Homepage”, https://mobility-dataspace.eu/de/ (accessed on 7 March 2025).
[140] Moumen, I., J. Abouchabaka and N. Raflia (2023), “Enhancing urban mobility: Integration of IoT road traffic data and artificial intelligence in smart city environment”, Indonesian Journal of Electrical Engineering and Computer Science, Vol. 32/2, pp. 985-993, https://pdfs.semanticscholar.org/0135/9ee140d21de32affe28e5ad9a50254686d7e.pdf.
[8] Muench, S. et al. (2022), Towards a Green and Digital Future – Key Requirements for Successful Twin Transitions in the European Union, Publications Office of the European Union, https://doi.org/10.2760/977331.
[100] NEHA (2024), “Training seminars on the use of the EAHY portal by health officers for the electronic cross-border health services”, Cyprus National eHealth Authority, https://www.neha.org.cy/en/training-seminars-on-the-use-of-the-eahy-portal-by-health-officers-for-the-electronic-cross-border-health-services-patient-summary-eprescription-edispensation/ (accessed on 16 February 2025).
[17] NetZeroCities (2024), “Slovenia’s pilot activity: UP-SCALE-urban pioneers - systemic change amid liveable environment”, https://netzerocities.eu/slovenias-pilot-activity-up-scale-urban-pioneers-systemic-change-amid-liveable-environments/ (accessed on 17 February 2025).
[21] NexSys (2022), “About NexSys”, https://www.nexsys-energy.ie/about/ (accessed on 9 August 2024).
[127] NGC (2024), “About the Danish national genome center”, https://www.eng.ngc.dk/about-the-danish-national-genome-center (accessed on 15 March 2025).
[181] NHL Stenden (2024), “Smart traffic analys and traffic safety (STATS)”, https://www.nhlstenden.com/onderzoek/lectoraten/computer-vision-data-science/STATS (accessed on 8 May 2025).
[44] NIJZ (2023), “Supporting health data access bodies in Slovenia – SI-SUD”, Nacionalni Inštitut za javno zdravje RS, https://nijz.si/nijz/javni-pozivi/javni-poziv-za-izbor-zunanjega-strokovnjaka-ki-bo-sodeloval-pri-izvedbi-evalvacije-projekta-v-okviru-pogodbe-supporting-health-data-access-bodies-in-slovenia-si-sud/ (accessed on 14 January 2025).
[143] Nikitas, A. et al. (2020), “Artificial intelligence, transport and the smart city: Definitions and dimensions of a new mobility era”, Sustainability, Vol. 12/7, p. 2789, https://doi.org/10.3390/su12072789.
[203] NL AIC (2024), “AI-hub North-Netherlands”, https://nlaic.com/en/ai-hub/ai-hub-north-netherlands/ (accessed on 3 May 2025).
[89] NOA (2024), “METEO/NOA Unit”, National Observatory of Athens, https://www.iersd.noa.gr/en/operational-services/meteo/ (accessed on 16 March 2025).
[119] Nordic Council of Ministers (2024), “Supporting the healthcare professionals’ work and data quality through e-health standards”, https://pub.norden.org/temanord2024-514/index.html (accessed on 18 April 2025).
[118] Nordic Council of Ministers (2019), e-Health Standardisation in the Nordic Countries, Nordic Council of Ministers, Copenhagen, https://norden.diva-portal.org/smash/get/diva2:1340369/FULLTEXT01.pdf.
[160] nxtAIM (2025), “Generative Methoden für Perzeption, Prädiktion und Planung”, [English translation], https://nxtaim.de/ (accessed on 4 March 2025).
[204] NXTGEN Hightech (2025), “NXTGEN Hightech”, https://nxtgenhightech.nl/en/ (accessed on 8 January 2025).
[130] OECD (2025), Governing with Artificial Intelligence: The State of Play and Way Forward in Core Government Functions, OECD Publishing, Paris, https://doi.org/10.1787/795de142-en.
[131] OECD (2024), “2023 OECD Digital Government Index: Results and key findings”, OECD Public Governance Policy Papers, No. 44, OECD Publishing, Paris, https://doi.org/10.1787/1a89ed5e-en.
[132] OECD (2024), Global Trends in Government Innovation 2024: Fostering Human-Centred Public Services, OECD Public Governance Reviews, OECD Publishing, Paris, https://doi.org/10.1787/c1bc19c3-en.
[188] OECD (2023), Policies for the Future of Farming and Food in the European Union, OECD Agriculture and Food Policy Reviews, OECD Publishing, Paris, https://doi.org/10.1787/32810cf6-en.
[4] OECD (2022), “Measuring the environmental impacts of artificial intelligence compute and applications: The AI footprint”, OECD Digital Economy Papers, No. 341, OECD Publishing, Paris, https://doi.org/10.1787/7babf571-en.
[43] OECD (2019), Health in the 21st Century: Putting Data to Work for Stronger Health Systems, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/e3b23f8e-en.
[74] Oxipit (2024), “Chestlink – World’s first autonomous radiology AI solution”, https://oxipit.ai/products/chestlink/ (accessed on 17 April 2025).
[146] Paiva, S. et al. (2021), “Enabling technologies for urban smart mobility: Recent trends, opportunities and challenge”, Sensors, Vol. 21/6, p. 2143, https://doi.org/10.3390/s21062143.
[187] Patricio, I. and R. Rieder (2018), “Computer vision and artificial intelligence in precision agriculture for grain crops: A systematic review”, Computers and Electronics in Agriculture, Vol. 153, pp. 69-81, https://doi.org/10.1016/j.compag.2018.08.001.
[139] PLANAPP (2021), “PLANAPP presentation”, Portuguese Centre for Planning and Evaluation of Public Policies, https://www.planapp.gov.pt/planapp-presentation/ (accessed on 15 March 2025).
[40] PNAV (2025), “Homepage”, Programa nacional de algoritmos verdes, https://algoritmosverdes.gob.es/es (accessed on 26 March 2025).
[209] Portugal INCoDe.2030 (2022), “AI Portugal 2030 – Portuguese national initiative on digital skills”, Portugal INCoDe.2030, https://www.incode2030.gov.pt/en/wp-content/uploads/2022/01/julho_incode_brochura.pdf (accessed on 8 March 2025).
[92] Portuguese Ministry of Health (2024), “SNS 24 digital services”, https://www.sns24.gov.pt/ (accessed on 8 November 2024).
[63] QUANTUM (2024), “QUANTUM – The health data quality label”, https://quantumproject.eu/ (accessed on 13 November 2024).
[81] Radiobotics (2024), “AI-powered solutions for ED and radiology”, https://radiobotics.com/ (accessed on 14 November 2024).
[22] REA (2021), “Mākslīgais intelekts nākamās paaudzes enerģētikai (I-NERGY)”, [English translation], Riga Energy Agency, https://rea.riga.lv/projekti/maksligais-intelekts-nakamas-paaudzes-energetikai-i-nergy/ (accessed on 3 March 2025).
[78] Regina Maria (2023), “Centrele de imagistica Regina Maria – pionier in domeniul aplicatiilor de inteligenta artificiala in imagistica medicala”, [English translation], https://www.reginamaria.ro/articole-medicale/centrele-de-imagistica-regina-maria-pionier-do-meniul-aplicatiilor-de-inteligenta-artificiala (accessed on 16 April 2025).
[185] Rejeb, A. et al. (2022), “Examining the interplay between artificial intelligence and the agri-food industry”, Artificial Intelligence in Agriculture, Vol. 6/21, https://doi.org/10.1016/j.aiia.2022.08.002.
[25] RenovAIte (2025), “Unterstützung der Renovierungs-Industrie mit KI”, [English translation], https://renovaite.eu/de/home-de/ (accessed on 26 March 2025).
[107] Riga TechGirls (2024), “Programs”, https://rigatechgirls.com/programs/ (accessed on 14 April 2025).
[2] Ruf, B. (2024), “Navigating the environmental impact of AI”, The AI Wonk, https://oecd.ai/en/wonk/navigating-the-environmental-impact-of-ai (accessed on 18 October 2024).
[26] Sadales tīkls (2025), Jaunas lielo datu vadītas virszemes elektrotīkla pirms vētras bojājumu prognozēšanas sistēmas (ST ODA) izstrāde efektīvākai elektroapgādes traucējumu novēršanai, [English translation], https://sadalestikls.lv/en/st-oda (accessed on 3 March 2025).
[212] Secretariatul General al Guvernului (2024), Raportare Pentru Baza de Date Electronică A Comisiei Europene Cesar Referitoare la Punerea În Aplicare a Recomandărilor Specifice de Țară 2019-2023, [English translation], Secretariatul General al Guvernului, https://sgg.gov.ro/1/wp-content/uploads/2024/03/ANEXA-31.pdf.
[184] SIA (2024), “BikeSafeAI”, https://www.sia-projecten.nl/project/bikesafeai (accessed on 14 March 2025).
[174] Silicon Economy (2025), “Homepage”, https://www.silicon-economy.com/english/ (accessed on 17 March 2025).
[87] Sirris (2021), “Smart & social home care”, https://www.sirris.be/fr/joint-project/smart-social-home-care-au-service-du-bien-etre-des-personnes-agees-leur-domicile (accessed on 14 October 2024).
[36] Slovenian Ministry of Digital Transformation (2023), Digital Public Services Strategy 2030 – Action Plan, Slovenian Ministry of Digital Transformation, http://vrs-3.vlada.si/MANDAT22/vladnagradiva.nsf/71d4985ffda5de89c12572c3003716c4/8e2cd3e7ab3a5ac7c1258a1c003e959d/$FILE/ANSDJS2030AN3.pdf.
[30] Slovenian Ministry of Economic Development and Technology (2022), “The government has adopted strategic baselines for promoting energy efficiency in the economy”, https://www.gov.si/novice/2022-05-16-vlada-je-sprejela-strateska-izhodisca-za-spodbujanje-energetske-ucinkovitosti-v-gospodarstvu/ (accessed on 18 July 2024).
[94] Slovenian Ministry of Health (2024), “Supporting the digital transformation of healthcare”, https://www.gov.si/zbirke/projekti-in-programi/podpora-digitalni-transformaciji-zdravstva/ (accessed on 8 March 2025).
[93] Slovenian Ministry of Health (2022), “Slovenija – e-zdravje za bolj zdravo družbo”, [English translation], https://www.gov.si/assets/ministrstva/MZ/DOKUMENTI/staro/Slovenija-E-zdravje-za-bolj-zdravo-druzbo-v2.pdf (accessed on 8 March 2025).
[154] Slovenian Ministry of Higher Education, Science and Innovation (2023), “Public call for the selection of operations “Support to strategic development and innovation partnerships (SRIPs) for the period 2023 – 2026””, GOV.si, https://www.gov.si/zbirke/javne-objave/javni-razpis-za-izbor-operacij-podpora-strateskim-razvojno-inovacijskim-partnerstvom-srip-za-obdobje-2023-2026/ (accessed on 7 February 2025).
[182] Slovenian Ministry of Infrastructure (2022), “National traffic management centre”, https://ncup.si/en/about (accessed on 8 February 2025).
[29] Smart Attica (2024), “AI-powering Greece: The #SmartAttica project”, https://www.smartattica.eu/ (accessed on 2 December 2024).
[163] Smart Dublin (2022), “Smart DCU: Delivery robots”, https://www.tii.ie/en/news/press-releases/eu-connected-vehicle-pilo/ (accessed on 16 March 2025).
[175] Smart Dublin (2020), “Mapping city assets from street level imagery”, https://smartdublin.ie/mapping-city-assets-from-street-level-imagery/ (accessed on 16 March 2025).
[16] SMASH (2024), “4. Climate – machine learning in climate research”, https://smash.ung.si/research-areas/4/climate/ (accessed on 17 March 2025).
[112] SNS (2024), “Academy SPMS”, https://academia.spms.min-saude.pt/ (accessed on 8 November 2024).
[102] SOLAS (2024), “Homepage”, https://www.solas.ie/ (accessed on 8 November 2024).
[47] SPMS (2024), “HealthData@PT”, Portuguese Ministry of Health, https://www.spms.min-saude.pt/healthdatapt-eng/about/ (accessed on 2 February 2025).
[52] SPMS (2023), “OurHealth@PT”, Portuguese Ministry of Health, https://www.spms.min-saude.pt/our_health_pt/ (accessed on 2 June 2024).
[101] SPMS (2023), “SPMS promotes the 1st webinar MyHealth@Europe”, https://www.spms.min-saude.pt/2023/12/spms-promove-1-o-webinar-a-minha-saude-europa/ (accessed on 2 June 2024).
[116] Statens Serum Institut (2025), EU-HIP, Statens Serum Institut.
[123] STM (2024), “EU collaboration project to promote access to genomic data”, Finnish Ministry of Social Affairs and Health, https://stm.fi/en/european-1-million-genomes-initiative (accessed on 8 April 2025).
[69] Stratégie de Spécialisation Intelligente Wallonne (2024), “Médecine des réseaux et des systèmes (MedResyst)”, https://s3.wallonie.be/home/etre-acteur-de-l-innovation/les-initiatives-d-innovation-strategiques/dis-2-sante-renforcee/medresyst.html (accessed on 7 February 2025).
[28] sustAIn.brussels (2024), “Homepage”, https://www.sustain.brussels/ (accessed on 3 December 2024).
[61] Swedish eHealth Agency (2024), “National work for interoperability in healthcare”, https://www.ehalsomyndigheten.se/om-oss/regeringsuppdrag/nationell-funktion-for-interoperabilitet/ (accessed on 7 April 2025).
[106] Swedish Government (2020), “Vision for eHealth 2025”, Swedish Government, https://ehalsa2025.se/wp-content/uploads/2021/02/Strategy-2020-2022_eng.pdf.
[168] TACR (2024), “Program DOPRAVA 2030”, Technology Agency of the Czech Republic, https://www.tacr.cz/program/doprava-2030/ (accessed on 13 March 2025).
[164] TEC (2022), “TEC à la demande et Autonom-e : bilan et perspectives”, https://presse.groupe.letec.be/tec-a-la-demande-et-autonom-e--bilan-et-perspectives (accessed on 3 December 2024).
[169] Technology Incubator (2024), “Mobility innovation hub”, https://technologickainkubace.org/huby/mobility-innovation-hub/ (accessed on 24 November 2024).
[113] THCS (2022), “Strategic research and innovation agenda”, European Partnership on Transforming Health and Care Systems, https://www.thcspartnership.eu/thcs/strateig-overview/strategic-research-and-innovation-agenda.kl (accessed on 16 February 2025).
[162] Transport Infrastructure Ireland (2025), “EU connected vehicle pilot study”, https://www.tii.ie/en/news/press-releases/eu-connected-vehicle-pilo/ (accessed on 3 February 2025).
[11] UBA (2024), “The AI lab at the German Environment Agency”, German Environment Agency, https://www.umweltbundesamt.de/en/topics/digitalisation/uba-application-lab-for-ai-big-data (accessed on 1 February 2025).
[34] University of Malta (2023), “AI employed in the safeguarding of the Megalithic temples of Malta – The MEGALITH project”, https://www.um.edu.mt/newspoint/news/2023/11/ai-employed-in-the-safeguarding-of-the-megalithic-temples-of-malta-the-megalith-project (accessed on 17 September 2024).
[111] VAIA (2024), “Flanders AI academy”, https://www.vaia.be/en/ (accessed on 8 November 2024).
[194] Vangeyte, J. (2025), “Precisielandbouw en datatechnologie”, [English translation], https://ilvo.vlaanderen.be/nl/themas/precisielandbouw-en-datatechnologie (accessed on 17 January 2025).
[215] VistaMilk (2025), “VistaMilk”, https://www.vistamilk.ie/ (accessed on 10 January 2025).
[18] VUB (2024), “CPC-FARI 2022: Coordination mechanisms for the sharing of energy through proxies, from the user to general guidelines”, Vrije Universiteit Brussel, https://researchportal.vub.be/en/projects/cpc-fari-2022-coordination-mechanisms-for-the-sharing-of-energy-t.
[37] VUB (2024), “Joint R&D 2022: REASSURE: Context-aware reliability assessment and optimisation of sustainable energy systems”, Vrije Universiteit Brussel, https://researchportal.vub.be/en/projects/joint-rd-2022-reassure-context-aware-reliability-assessment-and-o (accessed on 3 December 2024).
[150] Wang, X. et al. (2023), “DriveDreamer: Towards real-world-driven world models for autonomous driving”, arXiv preprint arXiv:2309.09777, https://doi.org/10.48550/arXiv.2309.09777.
[85] WEF (2024), “AI in healthcare: Buckle up for change, but read this before takeoff”, https://www.weforum.org/stories/2024/01/ai-in-healthcare-buckle-up-for-big-change-but-read-this-before-takeoff/ (accessed on 29 January 2025).
[218] Wu, C. et al. (2022), “Sustainable AI: Environmental implications, challenges and opportunities”, Proceedings of Machine Learning and Systems, Vol. 4, pp. 795-813, https://proceedings.mlsys.org/paper_files/paper/2022/hash/462211f67c7d858f663355eff93b745e-Abstract.html.
[207] WUR (2025), “AI in dier- en landsystemen”, [English translation], https://www.wur.nl/en/research-results/research-funded-by-the-ministry-of-lvvn/types-research/soorten-onderzoek/kennisonline/ai-in-dier-en-landsystemen.htm (accessed on 2 June 2025).
[58] Xt-EHR (2024), “Standardization for the adoption of the EEHRxF@EU”, Joint Action Extended EHR@EU Data Space for Primary Use, https://www.xt-ehr.eu/ (accessed on 13 November 2024).
[192] Zhou, J. and F. Chen (2023), Artificial Intelligence in Agriculture, Springer, Cham, https://doi.org/10.1007/978-3-030-89123-7_183-2.
← 1. Reflecting this trend, the OECD.AI Policy Observatory encompasses a dedicated space on AI in Government.