Sisendi küsimine

EN EN

EUROPEAN COMMISSION

Brussels, 19.11.2025

COM(2025) 835 final

COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN

PARLIAMENT AND THE COUNCIL

DATA UNION STRATEGY

UNLOCKING DATA FOR AI

1

1. Introduction - Unlocking data for artificial intelligence

Artificial intelligence is transforming the global economy, and, the EU needs large volumes of

high-quality data to compete and drive innovation. Without such data, the EU cannot build

strong AI models, optimise healthcare or the energy system, or sustain industrial leadership.

For small-and medium enterprises in particular, better data access will be decisive for scaling

and remaining competitive.

The EU has laid strong foundations for the creation of a secure, interoperable single market for

data through key legislations such as the Data Act1, and investing in common European data

spaces2. At the same time, the AI Continent Action Plan3 and the Apply AI Strategy4 have

created the conditions for the EU to lead in AI development and uptake.

Yet, the EU is facing a scarcity of data for AI development, and growing geopolitical

competition where data is increasingly seen as a strategic asset. Much valuable data remains

siloed or underused, also due to a complex patchwork of data rules, while global competitors

move faster to exploit it for technological and industrial advantage.

To facilitate compliance and improve predictability, the digital omnibus proposes to simplify

the data regulatory landscape by merging four legal instruments into one single, coherent

data framework. Moreover, to support companies and ease compliance, the strategy will be

accompanied by a comprehensive support package under the Data Act. Model contractual

terms standard cloud clauses and a dedicated helpdesk will help SMEs in particular navigate

obligations, reduce legal complexity and focus on innovation. Model clauses will apply to both

B2G and B2B relations, supporting data creation, sharing and simpler contracts.5

The Data Union strategy shifts the focus from rules to results, To achieve this, the EU will act

in three priority areas:

• scaling up access to data for AI, with initiatives such as data labs that offer trusted

pseudonymisation services and pool data resources across public and private actors to

provide companies and researchers with high-quality datasets.

• streamlining data rules to make sharing data easier for businesses and researchers,

including reforming cookie consent to reduce fatigue while protecting rights.

• strengthening the EU’s global position on international data flows, by tackling

unjustified trade barriers so that European companies can compete on a level playing

field globally.

1 Regulation (EU) 2023/2854 of the European Parliament and of the Council of 13 December 2023 on harmonised rules on

fair access to and use of data and amending Regulation (EU) 2017/2394 and Directive (EU) 2020/1828 2 European Commission, Commission Staff Working Document on Common European Data Spaces, SWD(2024) 21 final, 24

January 2024. 3 European Commission (2025). AI Continent Action Plan. Communication from the Commission to the European Parliament,

the Council, the European Economic and Social Committee and the Committee of the Regions. COM(2025) 165 final.

Brussels: European Commission. 4 European Commission, Apply AI Strategy, COM(2025) 723 final, Brussels, 8 October 2025 5 Updated EU AI model contractual clauses | Public Buyers Community

2

2. Building on the European strategy for data (2020–2025)

With the 2020 European strategy for data6, the EU created the legal and institutional

foundations for a secure and fair single market for data. The goal was to unlock the potential

of data for innovation and growth while protecting rights. However, with generative AI and

rising geopolitical competition, it is clear that the EU needs to go beyond the foundations it has

built.

The European strategy for data was the driver of key legislation to build trust, promote data

sharing, and clarify rules across the data value chain. The Data Governance Act created

mechanisms for trustworthy data sharing, regulated intermediaries, introduced a framework for

voluntary sharing of data by companies for purposes of general interest (voluntary data

altruism), and opened up certain protected public-sector datasets. The Data Act unlocks data

from connected products and services by clarifying access and usage rights. Lastly, under the

Open Data Directive and its Implementing Act on high-value datasets (applicable since June

2024) certain public sector datasets had to be made freely and openly available in machine-

readable formats. However, inconsistent national implementation and uncertainties around

trade secrets are some of the remaining challenges of the existing legislative framework.

Supporting measures that were set up under the European strategy for data include working

with the European Data Innovation Board to coordinate Member States’ efforts and a

standardisation request to lay the groundwork for a European trusted data framework.7

To make the European single market for data a reality,

between 2021 and 2024 the Commission also invested €336

million in 14 strategic common European data spaces

spanning key economic sectors and areas of public interest,

complementing national and private-sector efforts. These

spaces provide secure infrastructure and governance

frameworks for voluntary data sharing under agreed

conditions. The main challenge now is scaling up these

efforts for EU-wide impact.

3. Three challenges the EU must address now

As AI technology and services reshape the global landscape, the EU must urgently confront

three new, strategic challenges: data scarcity, regulatory complexity, and rising global

competition.

6 The European data strategy – Shaping Europe’s digital future, Publications Office,

2020, https://data.europa.eu/doi/10.2775/645928 7 European Commission, Commission Implementing Decision C(2025) 4135 of 1 July 2025 on a standardisation request to

the European standardisation organisations as regards a European Trusted Data Framework in support of Regulation (EU)

2023/2854 of the European Parliament and of the Council, available at: https://ec.europa.eu/growth/tools-

databases/enorm/mandate/614_en (accessed on 27 October 2025)

The European Cancer Image

Data Space covers

anonymized images and

annotations. By 2027, it will

include more than 60

million cancer images.

3

Data Scarcity: A structural bottleneck for innovation

With the rise of generative AI, large language models (LLMs) and Agentic AI8, access to high-

volume, high-quality, unseen, and domain-specific datasets has become a defining factor of

global competitiveness. According to Epoch AI, the size of datasets used to train LLMs doubles

approximately every six months.9

LLMs and other kinds of foundation models demand massive, diverse sets of training data.

Studies suggest that, at current trends, the volume of publicly available training data could be

exhausted between 2026 and 2032.10

The EU’s challenge is twofold: (i) to make high-quality datasets, including also sector-specific

datasets, more widely available; and (ii), to ensure the computing infrastructure needed to

process these datasets is accessible at scale. Many European firms, especially SMEs and start-

ups, lack the data volume and diversity and the access to European compute capacities needed

to develop competitive AI solutions. Without urgent action, the EU risks being left behind.

Regulatory complexity: fragmentation hampers scale

Following the 2020 European strategy for data, the EU introduced landmark regulations

building on pre-existing rules - the Data Governance Act11, the Data Act, and various sectoral

laws such as the European Health Data Space Regulation12. Each of these initiatives focused

on specific issues, such as mechanisms for data sharing, fair distribution of value and tackling

burdensome localisation requirements. However, complex interplay between the General Data

Protection Regulation (GDPR)13 and sectoral laws, and uneven implementation across Member

States created a fragmented regulatory landscape, legal uncertainty, including for public

authorities, and raise compliance costs, especially for start-ups and SMEs.

For example, providers of data intermediation services – still an emerging field – are subject

to restrictive legal obligations that limit their ability to grow. There is a need to avoid burdening

early-stage ecosystems with disproportionate requirements that impede the uptake of data-

sharing models and the roll-out of data spaces. To unlock innovation, the EU must simplify the

rules on data access and use.

8 “Agentic AI are AI systems that can independently make decisions and take actions. This enables agents to understand

language, reason about tasks, take actions autonomously to achieve predefined objectives, and interact with the world around

them, orchestrating interactions including with humans 9 Robi Rahman and David Owen (2024), "The size of datasets used to train language models doubles approximately every six

months". Published online at epoch.ai. Retrieved from: 'https://epoch.ai/data-insights/dataset-size-trend' [online resource] 10 Villalobos, P., Ho, A., Sevilla, J., Besiroglu, T., Heim, L., & Hobbhahn, M. (2024). Position: Will we run out of data? Limits

of LLM scaling based on human-generated data. In K. Chaudhuri, S. Jegelka, L. Song, D. L. Silver, & Y. Ermon (Eds.),

Proceedings of the 41st International Conference on Machine Learning (Vol. 235, pp. 42085–42101). PMLR.

https://proceedings.mlr.press/v235/villalobos24a.html 11 Regulation (EU) 2022/868 of the European Parliament and of the Council of 30 May 2022 on European data governance

and amending Regulation (EU) 2018/1724 (OJ L 152, 3.6.2022, p. 1) 12 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 13 Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of natural

persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC

4

Global competition: data as a strategic asset

In the AI race, access to high-value data is a key strategic advantage. Globally, data has become

a geopolitical asset, with data access, localisation, and control increasingly used as instruments

of power. While the EU promotes open, secure, fair and trusted data flows, other jurisdictions

follow assertive or protectionist strategies. Localisation and restrictive access regimes abroad

limit the EU’s access to global resources and expose EU firms to economic and security risks.

To unlock the full potential of European AI, the Union must treat data as a core strategic

resource and invest in secure, high-quality, and interoperable datasets that reflect European

values and standards. Strengthening Europe’s ability to collect, curate, and use its own data is

both an economic and security imperative. The EU must secure beneficial flows, safeguard

sensitive non-personal data within the EU, and support digital sovereignty amid intensifying

technological rivalry.

4. The three pillars of the Data Union Strategy

Pillar I: Scaling up access to quality data

for AI and innovation

The EU’s competitiveness in AI and digital

innovation depends on access to high-quality

data and the infrastructure to share and use

data securely at scale. The EU has already laid

strong foundations with common European

data spaces, governance frameworks, and

major investments in cloud technology and

computing. The challenge now is to move

from pilot projects and fragmented initiatives

to a seamless, interoperable, and sustainable

data ecosystem, encouraging breakthrough

innovation and strengthening the EU’s digital

sovereignty.

To achieve this, the Commission will act along

two complementary tracks. First, it will launch

flagship initiatives that address the EU’s most

immediate bottlenecks: limited access to

critical datasets, insufficient infrastructure for

large-scale AI development, and the need for

trusted environments, including data labs that

connect data spaces with AI developers. These

data labs will serve as specialised service

facilities providing secure environments,

practical tools, and expert support for data pooling, curation, pseudonymisation, and

Data Spaces and Data Labs: The Building

Blocks of Europe’s AI Ecosystem

Common European Data Spaces are data-

sharing ecosystems built on cloud

infrastructure and clear governance rules

defining who can access, use, and share data.

They connect public and private actors around

trusted mechanisms for data exchange within

and across sectors.

Data Labs are data service providers that

link these data spaces with the AI

ecosystem. They give companies and

researchers secure, practical access to

high-quality datasets, the support they need

to ensure compliance with EU rules, and

offer tools, guidance, and trusted

environments for data pooling, curation,

labelling and pseudonymisation.

Data spaces provide the structured sources

of trustworthy data, while data labs turn

this data into usable resources for

innovation and AI development, ensuring a

seamless flow from availability to

application.

5

anonymisation. They will help companies, especially SMEs, turn data into usable resources for

AI training while preserving data control. These efforts will work hand in hand with the Apply

AI Strategy, ensuring that data availability directly supports AI deployment and innovation

across industries and public sectors. Second, it will reinforce these efforts with horizontal

enablers: legal clarity for data pooling, standards for data quality, and investment in synthetic

data14 capacities, ensuring scale, trust, and long-term sustainability across all sectors.

i. Scaling up the common European data spaces

The common European data spaces (CEDS) are central to building a single market for data.

The next phase will scale them up and link them to AI infrastructure through data labs and AI

factories, turning the EU’s data assets into fuel for trustworthy AI. In close synergy with the

Apply AI Strategy, these efforts will ensure that data spaces directly enable AI development

and deployment across sectors.

Simpl cloud middleware15 will enable interoperability across initiatives through an open-

source, modular, and secure set of components. This lowers barriers for SMEs and creates

faster links between ecosystems. The data spaces support centre will reinforce uptake,

especially among SMEs, by raising awareness and practical guidance.

Future EU funding for CEDS will prioritise sectors of public interest, such as health, mobility,

energy, public administrations and the environment, while mature domains like manufacturing

and finance transition to market-driven

models. The Commission will support this

transition by promoting standardisation,

interoperability, and co-investment

frameworks. End-user integration, AI

readiness, and financial sustainability will

remain key objectives.

Among the flagship actions under the

Apply AI Strategy, the EU will leverage

Common European Data Spaces to

accelerate AI deployment across key

sectors and support the development of

Frontier AI models through the Frontier AI

Initiative. These actions are closely linked

to other Apply AI flagships, such as

Foundational Models for Industry, AI-

powered Pharma Discovery, Autonomous Drive Ambition Cities, each drawing on sectoral

14 Synthetic data is artificially generated data that is not collected from real-world events but is engineered to statistically

mimic the properties, patterns, and relationships of a real dataset. 15 Simpl is an open source, secure middleware that supports data access and interoperability in European data initiatives. It

provides multiple compatible components, free to use, that adhere to a common standard of data quality and data sharing;

https://simpl-programme.ec.europa.eu/

Next Steps for European Health Data Space:

The EHDS will serve as a key bridge between

health data ecosystems and AI development,

allowing data labs and AI Factories to

leverage anonymised and synthetic datasets

within trusted processing environments.

From March 2029, patient summaries and

ePrescriptions will be exchanged across all

Member States, alongside secondary use of

most health data. By March 2031, this will

extend to medical images, lab results and

discharge reports, with genomic and other

data added for secondary use.

6

data made available through the Common European Data Spaces. This approach translates into

concrete applications: AI-powered screening centres in healthcare that validate diagnostic tools

using the European Health Data Space16; trusted data pooling in manufacturing through the

Data Space for Manufacturing to train specialised and frontier AI models; and an Agri-Food

AI Platform that supports the uptake of AI-enabled farming tools using the Common European

Agricultural Data Space.

As of 2026, the roll-out of data spaces across priority sectors will continue, supported by

ongoing EU investment of around EUR 100 million enabling trusted and large-scale data use

for AI applications. The European Health Data Space will support AI-based diagnostics and

personalised medicine and serve as a key bridge between health data ecosystems and AI

development, allowing Data Labs and AI Factories to leverage anonymised and synthetic

datasets within trusted processing environments; the common European Mobility Data Space

will enable the connection vehicles, infrastructure, and logistics for safer, greener transport; the

Energy Data Space will facilitate smart and flexible energy services; and the Media Data

Space will boost creative industries through AI-driven cultural innovation. Data labs will act

as practical entry points to these data spaces, helping organisations access, prepare and use the

data effectively for AI. Within this framework, the European legal data space will expand

access to legal and judicial data through common identifiers and metadata for case law and

legislation, enabling LegalTech to use this data. The need for a contract terms data pool for

automated contracting will be explored in this context.

The Commission will fast-track environmental digitalisation through the Green Deal Data

Space, enabling the DigitalGreenTech community to scale cross-sector solutions using

reusable components and high-quality datasets. Priority actions include data-driven services

for the European Water Resilience Strategy, digitisation of permitting processes, pilots on

textile traceability and nature credits, and advanced forest monitoring with machine learning

on open and confidential data.

A European Defence Data Space will create a trusted environment for pooling operational,

industrial, and research data to develop next-generation defence systems, boost industrial

capabilities, and strengthen EU technological sovereignty by reducing reliance on third-

country providers. Drawing on Ukraine’s experience in data-driven defence, the Commission

will explore cooperation and knowledge exchange. The initiative will be developed with

Member States and relevant stakeholders, including businesses. 17

ii. Data labs

As outlined in the AI Continent Action Plan, data labs will be specialised facilities, linking data

holders, common European data spaces, domain-specific data ecosystems, and the EU AI

16It will also build on Europe’s Beating Cancer Plan, the Life Sciences Strategy, and the EU Cardiovascular Health Plan 17 This initiative will be guided by the European Defence Agency’s feasibility study due by end-2025

7

ecosystem. Data labs18 will provide hands-on services – such as data pooling19, curation20,

labelling and pseudonymisation21 – to help organisations, in particular start-ups and scale-ups,

share and use data safely, facilitate cooperative AI training and support the development of AI

models in key sectors and covering different governance and licensing models. In line with the

Apply AI Strategy, data labs will translate the availability of high-quality data into concrete AI

deployment, serving as practical enablers that accelerate experimentation, adoption and

scaling.They can also be used to carry out tasks that require advanced AI resources on behalf

of Data Spaces and other Data infrastructures, for example producing synthetic data, or

carrying out advance privacy and business secrecy preserving, to help organisations to share

and use data safely.

By pooling public and private resources, data labs will help overcome a key market failure:

limited availability of diverse, high-quality data and reluctance to share privately held data for

AI training. They will operate through existing access channels and frameworks without

requiring direct data transfer. In doing so, data spaces remain the trusted infrastructures where

data is governed and made available, while data labs can act as the operational interface that

enables its safe, value-adding use for AI.

Participation will be voluntary, and data holders decide how, when, and by whom data can be

used. No data will be transferred without explicit consent. All activities will be protected by

strict confidentiality safeguards and supported by privacy-preserving and decentralised

techniques such as federated learning, homomorphic encryption, and secure multi-party

computation. Data can be processed locally or across nodes without being merged into a single

repository, ensuring that it remains under the control of the original holder. This model –

particularly beneficial for SMEs – supports compliance with EU data protection rules,

safeguards confidentiality, and builds trust while expanding data use for AI.

The EU’s compute capacity has evolved from science-oriented high-performance computing

(HPC) under EuroHPC to AI Factories, which expand this concept to support AI development,

linking compute infrastructure with data access and experimentation. The upcoming AI

Gigafactories will further scale AI compute facilities.

Within this framework, the first data labs will be established under the AI Factories initiative

through EuroHPC, providing secure environments and data services to connect AI developers

with common European data spaces in areas such as healthcare, manufacturing, energy and

climate, and expanded to languages, cybersecurity, and cultural heritage. To ensure their

services reach companies and public administrations, data labs will work in close coordination

18 In some contexts, the term “data containers” is used to refer to similar facilities that enable structured, secure, and trusted

data use across different settings. Together with the broader concept of “data containerisation,” they reflect a complementary

approach to organising and governing data exchange, promoting interoperability and consistency across the EU AI ecosystem. 19 Combining and sharing of data from multiple sources into a single, centralized repository or shared environment. 20 Organising, integrating, validating, and maintaining data including its labelling for improving access and use. 21 Art 4(5) of the Regulation (EU) 2016/679: “The processing of personal data in such a manner that the personal data can no

longer be attributed to a specific data subject without the use of additional information, provided that such additional

information is kept separately and is subject to technical and organisational measures to ensure that the personal data are not

attributed to an identified or identifiable natural person.”

8

with the European Digital Innovation Hubs (EDIHs), which act as user-facing contact points

and help match data needs with concrete applications.

Further Data Labs will be set up independently in other domains to address specific sectoral or

research needs, such as the energy sector. The upcoming AI Gigafactories will further scale AI

compute facilities and prepare the Data Lab model for commercial rollout across the EU,

turning it into a self-sustaining service ecosystem that connects compute, data, and AI

innovation.

Data labs will provide services specifically across nine key areas:

• Bridge between data spaces and AI ecosystems: practical linkage that enables

companies to access high-quality, interoperable data by connecting common European

data spaces with AI developers, infrastructures and sectoral ecosystems.

• Technical infrastructure and tools: data containers will enable efficient storage and

organisation of data complemented by secure environments for the on-site processing

of sensitive data, along with ready-to-use tools for data preparation and privacy-

preserving techniques to achieve anonymisation and synthetic data generation. A high

standard of usability, speed, and scalability will be ensured so that tools are simple,

reliable, and easy to adopt.

• Data pooling: support for companies in aggregating data from public and restricted

sources - particularly data used for innovative purposes -, using the trusted data sharing

mechanisms of the common European data spaces. Data Labs will support businesses

to comply with EU competition law when exchanging or pooling data. Building on and

complementing the Horizontal Guidelines, which provide companies with practical

guidance on collaboration and shared resources, the Commission will further support

Data Labs in this role with dedicated guidance on best practices in data exchange and

pooling. In addition, tailored guidance for individual data labs following a request under

the Informal Guidance Notice will be available.

• Pseudonymisation and anonymisation services: provision of advanced tools and

expertise to remove or mask personal identifiers. These services will include techniques

such as pseudonymisation, anonymisation, and differential privacy, enabling safe data

reuse while maintaining analytical utility.

• Synthetic data generation: support for creating high-quality synthetic datasets that

replicate the statistical properties of real data without exposing sensitive or confidential

information. Data labs will provide tools and expertise to generate, validate, and

benchmark synthetic data for AI model training and testing, complementing

anonymisation efforts and improving data availability in sensitive domains.

• Data curation, labelling, and vectorisation: comprehensive support for cleaning,

labelling, annotating, enriching, and vectorising datasets to make them reliable,

representative, and usable for AI training. This includes quality assurance processes,

transparent documentation, and collaboration with expert communities for domain-

specific labelling.

9

• Regulatory guidance and training: tailored advice to help businesses comply with

EU law combined with training for AI developers on data use and legal obligations,

such as AI regulations, copyright, trade secrets and competition law, combined with

training for AI developers on data use and legal obligations.

• Bridge between data spaces and AI ecosystems: practical linkage that enables

companies to access high-quality, interoperable data by connecting common European

data spaces with AI developers, infrastructures and sectoral ecosystems.

Data access facilitation: a demand-driven service where start-ups and SMEs can signal

their data needs, with data labs helping them find relevant datasets and overcome

market, legal or administrative barriers.

iii. The Cloud and AI Development Act

Sustainable data centre capacity and sovereign cloud and AI services are a prerequisite for the

EU to achieve the objectives laid down in this strategy. As increasing amounts of data are

generated, there is a growing need to collect, store, combine and process this data. To minimise

How would a data lab work in practice?

A company in Member State X develops AI-based predictive maintenance systems for

electric vehicles but struggles to access enough high-quality sensor data from different car

models and charging infrastructures. Individual manufacturers are hesitant to share this

data due to trade secrets, privacy, and competition concerns. AI Factories will provide the

computing resources and, through their integrated data labs, data management services

needed to overcome these barriers.

Through the data lab, the company would access trusted, anonymised, and aggregated

datasets coming from different sources, such as public charging operators, participating

Original Equipment Manufacturers (OEMs) and other data discovered via the European

mobility data space.

As a component of the AI Factory, the data lab would offer:

• Secure environments to analyse real-time sensor data through federated learning

without the data leaving OEM systems.

• Anonymisation services ensuring privacy-compliant use of driver and vehicle data.

• Regulatory guidance on applying the Data Act’s data access provisions and

managing trade secret protection.

• Data curation tools that harmonise different sensor formats and quality standards.

The lab thus would act as a bridge between the mobility data space and the AI ecosystem,

allowing the company to train robust AI models while safeguarding manufacturers’

confidentiality.

10

latency22 and decrease reliance on infrastructure located in other parts of the world, the EU

needs to house sufficient data centre capacity.

To ensure the availability of sustainable data centre infrastructure and sovereign cloud and AI

services for EU businesses and public administrations, the Commission will propose a Cloud

and AI Development Act in Q1 2026. This initiative will support innovation across the entire

cloud and AI value chain, from the integration of cutting-edge processors to sustainable cooling

technologies and AI hardware and software. It will also accelerate the roll-out of sustainable

data centre capacity, ensuring that the EU has the infrastructure needed for secure and

sovereign cloud and AI services.

iv. Strategic data assets: public sector, scientific, cultural and linguistic resources

The EU’s competitiveness in AI depends on access to high-quality, structured, and trustworthy

data. Scientific, cultural, and linguistic datasets are critical enablers for robust AI models,

research breakthroughs, and technological sovereignty.

Public-sector reference datasets under the Open Data Directive will be scaled up. The high-

value datasets23 have to be made available free of charge, through application programming

interfaces (APIs), in a machine-readable format and, where relevant, provided as a bulk

download. In 2026, the Commission will propose to expand the list of high-value datasets to

cover legal, judicial, administrative and other data. This will be beneficial for start-ups and

SMEs. The Commission will also monitor whether further datasets should be added.

Scientific data has already proven transformative, as seen with AlphaFold.24 Well-structured

databases reduce research and development (R&D) costs, accelerate innovation, and open up

new frontiers in materials, pharmaceuticals, energy, and biotech. To build on this, the

Commission will continue to map existing databases, to set priorities with experts, secure usage

rights, and fund new digital infrastructures according to the European strategy on research and

technology infrastructure. In this regard, the European Open Science Cloud (EOSC), the

common European data space for R&D, is developing a federation of data repositories with a

trusted platform for sharing and reusing high-quality, findable, accessible, interoperable and

reusable (FAIR) research data, tools and services across disciplines and borders in Europe. This

will support the scientific activities with AI in RAISE.25 In parallel, the forthcoming proposal

22 Latency is the time it takes for data to pass from one point of a network to another. 23 In line with Annex I of the Open Data Directive these high value datasets come from the following categories: geospatial,

earth observation and environment; meteorological; statistics, companies and company ownership, mobility. New categories

can be added. 24 AlphaFold is an Artificial Intelligence system developed by Deep Mind, which uses deep learning and large amounts of data

to predict protein structures. This helps accelerating breakthrough research in many fields of biology. 25 European Commission (2025). Communication from the Commission to the European Parliament and the Council – A

European Strategy for Artificial Intelligence in Science: Paving the way for the Resource for AI Science in Europe (RAISE).

Brussels, 8 October 2025, COM(2025) 724 final

11

for a European Research Area (ERA) Act26 will strengthen legal conditions to share, access

and reuse publicly funded research results, publications and data for scientific purposes.

The EU’s cultural and linguistic resources will also be scaled up. More than 30 million digitised

works from Europe’s cultural institutions will be made available for AI development, building

on the Europeana initiative.27 The Commission will explore how to strengthen cooperation and

encourage licensing between public broadcasters and AI providers, in order to make their

audiovisual archives accessible for AI training, taking into account the remuneration of

rightsholders.

Pilot projects under the European common language data space and the Alliance for

Language Technologies (ALT-EDIC) will crowdsource domain-specific datasets including

from smaller languages, adding to the 477 billion tokens already available - comparable to

leading AI training datasets. This will also help to ensure that rare languages are included in

AI Large Language Models (LLM) development, which will have an impact on the quality of

the results of AI systems in these languages.

v. Horizontal enablers: synthetic data, data pooling, and standards

Alongside flagship initiatives, the EU also needs horizontal measures that cut across sectors

and give scale to the entire data economy.

Synthetic data as a driver of AI leadership

Synthetic data28 can unlock AI training in areas where data is scarce or sensitive, from rare

disease research through to robotics or autonomous driving edge cases. It enables AI model

development without exposing personal or proprietary information, strengthening both

competitiveness and privacy-preserving innovation.

To harness this potential, the Commission will develop guidance and standards for trusted

synthetic data use, examine the related legal questions, consult on a voluntary European

certification scheme, and explore the possibility of setting-up a ‘synthetic data factory’ to

provide access to high-performance computing for large-scale dataset generation. Horizon

Europe will also fund cutting-edge R&D in synthetic data generation techniques.

26 European Commission, Forthcoming proposal for a European Research Area (ERA) Act, announced in the Commission

Work Programme 2025, Brussels, 11 February 2025, available at: https://commission.europa.eu/strategy-and-policy/strategy-

documents/commission-work-programme/commission-work-programme-2025 27 Europeana, The European digital platform for cultural heritage, available at: https://www.europeana.eu/en (accessed on 27

October 2025) 28 See definition above

12

Clearing the path for strategic data pooling

Many companies for example in health, mobility, energy, agriculture, and manufacturing lack

the large, diverse datasets needed to train advanced AI models. Pooling of data related to early

stages of the production cycle of products and services

could unlock shared benefits, but legal uncertainty and

fear of breaching competition law hold back collaboration.

The Commission will continue to act to provide legal

clarity for companies, in line with the call to turn rules into

results in the report on the future of European

competitiveness by Mario Draghi. The 2023 Horizontal

Guidelines on cooperation agreements between

competitors already explain when data pooling is

compatible with EU competition law, with practical

examples and safeguards.

To further facilitate lawful and effective data

collaboration through Data Labs, the Commission will

issue dedicated guidance on best practices in data exchange and pooling.

In addition, competition law guidance can be provided by the Commission upon request under

the Informal Guidance Notice for specific data-related multi-country projects and initiatives

that foster cross-border innovation, industrial resilience, and AI development. By making data

pooling a trusted and legally secure option, the EU can unlock efficiencies and accelerate

breakthroughs in key sectors.

Raising the bar on data quality and data capturing

Without reliable standards, even the most ambitious data-sharing efforts risk fragmentation and

low uptake. The European trusted data framework29 already sets rules for sharing, metadata,

and governance, but further work is needed to address emerging issues.

The Commission will launch a standardisation request for a European data quality standard

covering completeness, consistency, provenance, semantic clarity, and governance, giving

businesses, regulators, and researchers shared benchmarks for reliable datasets. This work will

complement the ongoing standardisation efforts on data quality and documentation under the

AI Act, ensuring coherence between data management and AI development requirements.

A dedicated initiative will aim to standardise annotation and labelling practices, making data

easier to find, combine, and reuse while ensuring trust in its origin and conditions of use, which

is critical for scaling AI training and cross-sector reuse. A multi-stakeholder workshop will

also investigate standards for data capture from connected products, sensors, and cameras -

29 See also European Commission, Implementing Decision C(2025) 4135 on the European Trusted Data Framework

Draghi report: “In particular,

to overcome the EU’s lack of

large data sets, model training

should be fed with data freely

contributed by multiple EU

companies within a certain

sector. It should be supported

within open-source

frameworks, safeguarded from

antitrust enforcement by

competition authorities.”

13

including sampling, metadata, timestamping, calibration, and integrity - addressing a key

barrier to effective data pooling and reuse.

Flagship actions

• Launch first data labs to scale data availability and link to AI ecosystems (Q4 2025). They

will also offer trusted pseudonymisation services.

• Launching the quality data for AI initiative: expanding high-value datasets under the

Open Data Directive (Q4 2026); setting up a stakeholder forum with public broadcasters

and AI developers (Q2 2026); making 30 million digitised cultural objects available for AI

training (Q4 2026); and launching a crowdsourcing initiative for domain-specific data and

language data in smaller European languages (Q2 2026).

Pillar II: Streamlining data rules

The EU’s data framework must remain clear, practical, and innovation friendly. To reduce

burdens and boost competitiveness, the Commission is presenting a legislative proposal,

known as the Digital Omnibus, aiming, amongst other things, to modernise and consolidate

the EU’s horizontal data acquis. In addition, the Commission will also announce work on one-

click compliance to enable automated regulatory reporting, and a support package for the Data

Act, including model contracts, standard clauses, guidance on compensation and trade secrets,

and a legal helpdesk for SMEs.

i. Simplifying the EU’s data acquis

The EU’s regulatory data framework has grown rapidly, creating new rights but also increasing

complexity and fragmentation. Simplification is needed to reduce compliance costs, make rules

easier to apply, and better support innovation.

To this end, the Commission is presenting the above-mentioned Digital Omnibus. It will update

the acquis, removing unnecessary burdens while safeguarding the core principles of the EU’s

data economy. The Omnibus will focus on the following priority reforms:

• Deleting outdated rules. The Omnibus will repeal the Free Flow of Non-Personal Data

Regulation,30 as its functions are already covered by the Data Act, while explicitly

preserving the principle of free movement of non-personal data and the ban on

unjustified localisation.

• Streamlining data-sharing rules. The Omnibus will repeal the Data Governance Act

(DGA) and migrate its essential provisions into the Data Act. Obligations for data

intermediaries will be clearer, lighter and voluntary to enable viable models and wider

uptake.

• Consolidating public-sector data-sharing. Rules now split between the DGA and the

Open Data Directive will be kept and merged into one Data Act chapter. This simplifies

obligations while preserving openness, transparency, and fair access. The new

30 Regulation (EU) 2018/1807 of the European Parliament and of the Council of 14 November 2018 on a framework for the

free flow of non-personal data in the European Union

14

framework will furthermore tackle power imbalances in data sharing, ensuring fair

conditions and tangible benefits for SMEs. Data labs will flag promising new public

sector datasets not yet covered.

• Modernising rules for cookies and similar

technologies. The Omnibus will reform the

rules on cookies currently in the ePrivacy

Directive and bring them into the GDPR

framework. It will propose practical solutions:

cookies and similar technologies for certain

low-risk purposes should be considered

lawful, while other purposes, the operators

should rely on one of the legal bases under the

GDPR. It will also simplify banners with one-

click options. It will oblige websites to respect

users’ preferences, also through their browsers. Beyond the Digital Omnibus, the

ePrivacy framework will be reformed to ensure that current rules meet today’s needs

and allow for effective protection of people and business,without compromising

fundamental rights and preserving independent journalism. The relevant provisions will

be integrated into other legal instruments, allowing the Directive to be ultimately

withdrawn.

• Developing an innovation-friendly privacy framework. Targeted GDPR

amendments will, in particular, clarify the notion of personal data, harmonise at EU

level when data protection impact assessments should be conducted, simplify data

breach notifications to supervisory authorities, streamline breach notifications via a

single EU entry-point, simplify information obligations where there are reasonable

grounds to expect that individuals already have the information and the risk to the data

subject is low; clarify that legitimate interest can be a legal basis for training AI,

including the incidental processing of special categories of data; clarify the provisions

on automated individual decision-making.

One key change concerns liberating data for AI through trusted anonymisation. Today,

uncertainty about sufficient anonymisation of personal data is a core concern, often

discouraging data sharing. Businesses struggle in particular to determine when

pseudonymised data no longer constitutes personal data for certain entities. This

uncertainty makes data sharing more complex where the GDPR requirements are

complied with out of precaution. The Commission will support businesses by

specifying the means and criteria to determine whether data resulting from

pseudonymisation constitutes personal data for certain entities.

This will include an assessment of the state of the art of available techniques and the

development of criteria to assess the risk of re-identification. While businesses remain

fully responsible for compliance with the GDPR, they can use the implementation of

those means and criteria to demonstrate that data cannot lead to the reidentification.

The amendments will also facilitate AI model training, with the appropriate safeguards.

Participant in the Data Union

public consultation: “ePrivacy

rules need to be urgently updated.

The rules set in place were designed

against a totally different

technological background and do

not reflect the current market

needs.”

15

The goal of these changes is to provide legal clarity for AI development, including cases

of incidental processing of sensitive data where developers have made genuine efforts

to remove such data. while protecting individuals’ rights and competitiveness of

businesses.

• Refining the Data Act for practical implementation. The essential features of the

Data Act will remain unchanged. At the same time, business-to-government data

sharing will be limited to emergencies, easing burdens while safeguarding crisis

response. Targeted additional adjustments will prevent data ‘leakage’ to outside the EU,

introduce tailored regimes for custom-made cloud services and remove the provisions

on smart contracts.

• Reducing burdens for scaling companies. A new category of small mid-caps (250–

749 employees) will extend SME-type provisions under the Data Act, the Open Data

Directive, and integrated DGA rules.

ii. Building a future-proof data framework

As part of the Digital Fitness Check, the Commission will continue reviewing the EU’s data

acquis to keep it coherent, proportionate, and innovation friendly. With particular attention to

SMEs, it will identify overlaps, gaps, and unclear interactions, including with sectoral data

laws, to create a more predictable cross-sectoral framework.

In addition, we will modernise digital legislation and data protection.31 Targeted adjustments

may ease compliance and strenghten enforcement, supporting the development of robust and

trustworthy innovations.

Data brokerage has become a growing concern, with certain companies collecting, aggregating,

and trading personal data without individuals’ awareness, meaningful consent, or control. Such

opaque practices undermine core principles of data protection law, privacy, distort competition,

and erode public trust in digital markets. A strengthened enforcement of the existing rules is

needed. The Commission will assess whether additional safeguards are needed to curb these

practices, enhance transparency in data trading, and ensure that individuals and businesses can

trust how data is accessed and exchanged across the Union.

iii. One-click compliance

Today, companies spend significant time and money on compliance. Even data already in

digital form must often be reformatted and resubmitted to multiple authorities, where it is

checked manually. This duplication creates fragmented oversight and diverts resources from

innovation.

Beyond simplifying rules, the EU is investing in technologies to automate compliance.

Through Horizon Europe and the Digital Europe Programme, it supports common data models,

interoperability frameworks, and automated analysis. Pilot projects already show how real-

31 Commission work programme EUR-Lex - 52025DC0870 - EN - EUR-Lex.

16

time, automated compliance checks can work in practice. The Digital Product Passport (DPP)

is an early example of this approach in product legislation.

Building on these experiences, “one-click compliance” would make regulatory requirements

machine-verifiable, turning company data into standardised digital compliance certificates -

much like the DPP enables automatic product compliance.

One-click compliance could be particularly valuable in areas like cybersecurity, where

companies face requirements under NIS232, the Cyber Resilience Act33, and other frameworks.

The European Business Wallet Regulation will be a key enabler of this approach. It will

provide a trusted and interoperable digital environment for storing, managing, and sharing

verifiable credentials, including compliance certificates. Companies could use European

business wallets to digitally identify themselves, identify and validate users of the ecosystem

and demonstrate conformity with multiple EU rules through the submission of compliance

certificates, while public sector bodies regulators are provided with secure, immediate access

to validated information. Over time, the European business wallet will become a common

infrastructure supporting administrative processes such as licensing, public procurement, and

access to funding, enabling seamless digital interactions between businesses and authorities

across the Single Market.

Determining who is accountable in case of errors, misuse, or system failures - whether the

company, the certifier, or the regulator - will be essential to ensure trust and legal certainty.

The Commission will therefore explore these issues in an upcoming public consultation,

assessing both the opportunities and the safeguards needed to build a reliable and accountable

automated compliance ecosystem.

Beyond cutting costs for SMEs and mid-caps, such a system would also give policymakers

insights into how rules work in practice, strengthening evidence-based regulation. One-click

compliance could become a cornerstone of the EU’s digital simplification agenda, aligning

competitiveness with trust and accountability.

iv. Helping businesses comply with the Data Act

The Data Act constitutes the key set of rules for using and sharing data. To ensure that

companies, especially SMEs and small mid-caps, can fully use its potential and focus on

innovation rather than red tape, the Commission has already issued a FAQ document34 and

32 Directive (EU) 2022/2555 of the European Parliament and of the Council of 14 December 2022 on measures for a high

common level of cybersecurity across the Union, amending Regulation (EU) No 910/2014 and Directive (EU) 2018/1972, and

repealing Directive (EU) 2016/1148 (NIS 2 Directive), OJ L 333, 27.12.2022, p. 80–152 33 Regulation (EU) 2024/2847 of the European Parliament and of the Council of 23 October 2024 on horizontal

cybersecurity requirements for products with digital elements and amending Regulations (EU) No 168/2013 and (EU) No

2019/1020 and Directive (EU) 2020/1828 (Cyber Resilience Act), OJ L [2847], 20 November 2024 34 European Commission, Frequently Asked Questions – Data Act, version 1.3, Brussels, 12 September 2025, available at:

https://digital-strategy.ec.europa.eu/en/library/commission-publishes-frequently-asked-questions-about-data-act

(accessed on 27 October 2025)

17

guidance on in-vehicle data35, and will further complement these with a broader package of

support measures.

Immediate measures include:

• model contractual terms for data sharing to reduce legal complexity, cut transaction

costs, and give businesses confidence when entering into new partnerships;

• standard contractual clauses for cloud services to make switching easier and contracts

fairer, supporting competition and innovation in the European cloud market.

Further measures, to be phased in, will include:

• Guidelines on reasonable compensation to clarify what can be charged for data sharing,

providing legal certainty to both data holders and data recipients (Q1 2026);

• New guidance on selected definitions in the Data Act (Q1 2026);

• A Data Act legal helpdesk to provide direct assistance for companies with concrete

questions on how to apply the new rules, giving priority to SMEs to ensure their queries

are addressed swiftly and with dedicated attention (Q4 2025).

Together, these measures will make the Data Act easier to navigate, reduce unnecessary costs,

and give companies the clarity and confidence they need to seize new opportunities in the EU’s

data economy. The Commission will closely monitor the uptake of the contractual tools, in

particular the model contractual terms and standard contractual clauses, and will review,

complement or adapt them as needed in line with international developments in data sharing.

The Commission will seek synergies between the public buyers community and the European

Data Spaces to enhance public-sector efficiency, drawing on the blueprint established between

the European Health Data Space and the Big Buyers Working Group on Healthcare

Efficiency.36

Flagship actions

• Proposal to consolidate data legislation (Q4 2025)

• Proposal to update ePrivacy rules on cookies and similar technologies (Q4 2025)

• Proposal for targeted GDPR adjustments (Q4 2025)

• Launching a one-click compliance initiative (from Q4 2025 onwards)

• Rolling out support measures for the implementation of the Data Act (from Q4

2025 onwards)

35 European Commission, Guidance on vehicle data, accompanying Regulation (EU) 2023/2854 (Data Act), C(2025) 6119

final, Brussels, 12 September 2025 36Can the European Health Data Space enable better procurement? – Big Buyers are investigating | Public Buyers

Community

18

Pillar III: Safeguarding the EU’s data sovereignty through a strategic international data

policy

Data sovereignty is at the core of the EU’s digital future. It means that the EU must retain

control over how data is accessed, used, and protected – both within its territory and abroad.

Sovereignty requires openness to trusted partners, including exchange of data across borders,

but on terms that are fair, secure, and consistent with EU values and interests. A situation in

which foreign actors enjoy unfettered access to the EU market while European companies face

unjustified barriers abroad cannot be sustained.

Safeguarding sovereignty also means protecting the EU’s resilience. Cyberattacks, technology

leakage, surveillance, and coercive dependencies put critical data at risk. The EU must ensure

the availability, integrity, and security of sensitive datasets, preventing their misuse or

exploitation, in particular by actors outside the EU.

To this end, the Commission will pursue a strategy that

combines openness with strength: making fair conditions for

data access and cross-border transfer a pillar of digital trade,

protecting sensitive EU non-personal data through clear

safeguards, and deepening cooperation with trusted partners. It

will also work to shape global governance models that reflect

EU interests and values and prevent fragmentation into rival

spheres. This strategy will complement the long-lasting EU

approach to safe personal data flows developed through the EU data protection acquis.

While the EU has built a robust legal framework and promoted “data free flow with trust”

internationally, new unjustified data localisation requirements, export controls, and

discriminatory rules abroad threaten to undermine sovereignty. The Commission will therefore

act more assertively to defend EU interests and regulatory autonomy, with proportionate

measures where openness is abused or vulnerabilities weaponised.

i. Fair cross-border data flows and safeguards for EU sensitive non-personal data

The Commission will embed fair conditions and effective control of cross-border data flows

into international digital trade. Structured exchanges, e.g. in the framework of the EU’s digital

partnerships and dialogues will address existing imbalances where EU data flows abroad

without adequate safeguards.

If gaps persist, and on the basis of objective criteria, the Commission will take proportionate

action in full respect of the Union’s international commitments. It will issue guidelines in Q2-

2026 to assess the treatment of EU entities by third countries and develop an anti-data-leakage

toolbox in Q1 2026 to address localisation demands, market exclusion, or insufficient

safeguards or any other unjustified treatment. This toolbox may draw on or be inspired by

In a stakeholder poll, 75%

of participants supported

a more assertive EU

approach to international

non-personal data flows.

19

instruments such as the Trade Enforcement Regulation37, the Anti-Coercion Instrument38, and

economic security considerations, as applicable, and will focus on technologies and best

practices to strengthen the EU’s resilience. Should structural distortions or persistent

discriminatory practices remain unaddressed, the Commission will, where necessary, consider

additional measures to ensure fair conditions for data access and use.

In parallel, the Commission will better protect EU sensitive non-personal data, complementing

the protection of personal data guaranteed through the GDPR and adequacy decisions. Working

with stakeholders, and following the results of in-depth risk assessments, it will adopt a first

package of targeted measures by Q3 2026.

ii. Linking EU data-sharing ecosystems with those of like-minded third countries

The EU’s legal framework for data protection, cybersecurity, enforcement, and judicial redress

is a reliable basis for foreign data holders. The Commission will foster secure, convergent and

interoperable links between EU data ecosystems and those of like-minded partners to attract

more data flows to the EU.

Planned measures include (i) supporting services and infrastructure such as the CEDS to enable

seamless cross-border sharing; (ii) providing tools like standard contractual clauses to help

businesses ensure lawful exchanges; (iii) and embedding commitments on cross-border data

sharing in bilateral and plurilateral international agreements.

To strengthen convergence and interoperability, the Commission will promote the European

Trusted Data Framework in international dialogues and the Digital Partnership Network. It will

also explore creating a trust label, potentially linked to the data spaces maturity model – a

standardised framework designed to assess the capabilities of data space initiatives - to support

cooperation with governments and businesses abroad.

iii. Boosting the EU’s voice in global data governance

Competing models of data governance are fragmenting the global landscape. The Commission

will intensify the promotion of EU approaches internationally, in particular in emerging

frameworks, and strengthen coalitions with like-minded partners.

By 2026, in line with the International Digital Strategy39, the Commission and the European

External Action Service (EEAS) will deepen and connect digital partnerships on data

governance, aligning with partners that share common objectives and further develop digital

trade agreements and digital chapters within traditional trade agreements. It will continue to

37 Regulation (EU) No 654/2014 of the European Parliament and of the Council of 15 May 2014 concerning the exercise of

the Union’s rights for the application and enforcement of international trade rules and amending Council Regulation (EC) No

3286/94, OJ L 189, 27 June 2014, p. 50–58 38 Regulation (EU) 2023/2675 of the European Parliament and of the Council of 22 November 2023 on the protection of the

Union and its Member States from economic coercion by third countries (Anti-Coercion Instrument), OJ L 322, 27 November

2023 39 European Commission and High Representative of the Union for Foreign Affairs and Security Policy, Joint Communication

to the European Parliament and the Council — An International Digital Strategy for the European Union, JOIN(2025) 140

final, Brussels, 5 June 2025

20

engage actively in fora such as the G7, the G20, the OECD and the UN, using instruments like

the OECD ‘Declaration on Government Access to Personal Data.’

Particular attention will be paid to promoting EU approaches and mutually beneficial

collaboration with candidate countries, potential candidate and closest neighbours. The EU

will also work with partners to explore setting up a shared platform for selected high-value

public data (e.g. cultural heritage) and pursue trusted arrangements on sensitive data flows,

government access, and sector-specific rules. a shared platform for selected high-value public

data (e.g. cultural heritage) and pursue trusted arrangements on sensitive data flows,

government access, and sector-specific rules.

Flagship actions

• Issuing guidelines to assess fair treatment of EU data abroad (Q2 2026)

• Creating a toolbox to counter unjustified localisation, exclusion, weak safeguards,

and data leakage (Q2 2026) and adopting measures to protect sensitive non-

personal data (Q3 2026)

5. The Data Union strategy: unlocking data for AI

To ensure competitiveness in the age of AI, the Data Union strategy shifts gears from setting

rules to delivering results. Building on the foundations in place since 2020, it tackles data

scarcity, regulatory complexity, and global competition.

The European Data Innovation Board will remain the central governance forum, reformed for

deeper technical debates and strategic dialogue with Member States and industry. In parallel,

the Apply AI Alliance will become the main channel for sectoral feedback, ensuring that

companies, researchers, and public actors shape implementation. The AI Observatory will track

emerging trends and translate them into policy insights.

Targeted actions will scale up high-quality data, simplify the regulatory landscape, and

strengthen the EU’s role in global data flows. For SMEs and innovators, this means cheaper

compliance, easier access to data, and a more conducive international environment.

Only what gets measured, gets done. This is why the Commission has announced a single

market roadmap to increase the pace and speed up the processes. The Data Union strategy can

contribute as appropriate to the roadmap to help guide policymakers and industry, in particular

SMEs, in removing barriers and completing the single market for data.

Working hand in hand with the Apply AI Strategy, the Data Union Strategy ensures that the

EU’s data foundations directly power the development, deployment, and uptake of AI across

all sectors.

The long-term vision is clear: a sovereign European data economy where data flows securely

and responsibly, powering AI, fuelling innovation, and reinforcing competitiveness.

    Resolutsiooni liik: Riigikantselei resolutsioon Viide: Justiits- ja Digiministeerium /  / ; Riigikantselei /  / 2-5/25-02368

Resolutsiooni teema: Euroopa Komisjoni teatised ELi andmeliidu strateegia ja tehisaru rakendamise strateegia kohta

Adressaat: Justiits- ja Digiministeerium Ülesanne: Tulenevalt Riigikogu kodu- ja töökorra seaduse § 152` lg 1 p 2 ning Vabariigi Valitsuse reglemendi § 3 lg 4 palun valmistada ette Vabariigi Valitsuse seisukoha ja otsuse eelnõu järgneva algatuste kohta, kaasates seejuures olulisi huvigruppe ja osapooli:

- COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT AND THE COUNCIL DATA UNION STRATEGY UNLOCKING DATA FOR AI, COM(2025)835

-COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT AND THE COUNCIL Apply AI Strategy,COM(2025)723

EISi toimiku nr: 25-0593 Tähtaeg: 30.01.2026

Adressaat: Haridus- ja Teadusministeerium, Kaitseministeerium, Kliimaministeerium, Kultuuriministeerium, Majandus- ja Kommunikatsiooniministeerium, Rahandusministeerium, Regionaal- ja Põllumajandusministeerium, Siseministeerium, Sotsiaalministeerium, Välisministeerium Ülesanne: Palun esitada oma sisend Justiits- ja Digiministeeriumile seisukohtade kujundamiseks antud eelnõude kohta (eelnõude infosüsteemi (EIS) kaudu). Tähtaeg: 21.01.2026

Lisainfo: Eelnõusi on kavas arutada valitsuse 12.02.2026 istungil ja Vabariigi Valitsuse reglemendi § 6 lg 6 kohaselt sellele eelneval nädalal (04.02.2026) EL koordinatsioonikogus. Esialgsed materjalid EL koordinatsioonikoguks palume esitada hiljemalt 31.01.2026.

Kinnitaja: Merli Vahar, Euroopa Liidu asjade direktori asetäitja Kinnitamise kuupäev: 19.12.2025 Resolutsiooni koostaja: Sandra Metste [email protected],  

.

Euroopa Komisjoni teatised ELi andmeliidu strateegia (COM(2025)835) ja tehisaru rakendamise strateegia kohta

(COM(2025)723)

Otsuse ettepanek koordinatsioonikogule

Kujundada seisukoht

Kaasvastutaja sisendi tähtpäev 21.01.2026

KOKi esitamise tähtpäev 04.02.2026

VV esitamise tähtpäev 12.02.2026

Peavastutaja Justiits- ja Digiministeerium.

kaasvastutajad Majandus- ja kommunikatsiooniministeerium, Rahandusministeerium, Kliimaministeerium, Sotsiaalministeerium, Välisministeerium, Kaitseministeerium, Kultuuriministeerium, Haridus- ja teadusministeerium, Regionaal- ja põllumajandusministeerium, Siseministeerium.

Seisukoha valitsusse toomise alus ja põhjendus

Algatuse vastuvõtmisega kaasneks oluline majanduslik või sotsiaalne mõju (RKKTS § 152¹ lg 1 p 2);

Sisukokkuvõte

1) ELi andmeliidu strateegia (Data Union Strategy) seab  eesmärgiks konkurentsivõimelise andmete ökosüsteemi loomise, mis toetab tehisaru arendamist ning tugevdab ELi tehnoloogilist konkurentsivõimet. Strateegia keskendub tehisaru arendamiseks vajalike andmete kättesaadavuse parandamisele, õigusraamistiku lihtsustamisele ning rahvusvahelistele andmevoogudele. Strateegias määratakse kindlaks kolm prioriteetset tegevusvaldkonda:

1. Tehisaru andmetele juurdepääsu laiendamine, et tagada ettevõtetele juurdepääs innovatsiooniks vajalikele kvaliteetsetele andmetele.

2. Andme-eeskirjade ühtlustamine, et tagada ettevõtjatele õiguskindlus ja vähendada nõuete täitmisega seotud kulusid.

3. ELi andmesuveräänsuse kaitsmine, et tugevdada ülemaailmset positsiooni rahvusvahelistes andmevoogudes.

2) Tehisaru rakendamise strateegia (Apply AI Strategy) on ELi üldine tehisaru käsitlev valdkondlik strateegia, mille eesmärk on suurendada strateegiliste sektorite

2

konkurentsivõimet ja tugevdada ELi tehnoloogilist suveräänsust. Selle eesmärk on edendada tehisaru kasutuselevõttu ja innovatsiooni kogu Euroopas, eelkõige väikeste ja keskmise suurusega ettevõtjate (VKEd) seas. Strateegias julgustatakse rakendama tehisaru esmast poliitikat, mille puhul tehisaru peetakse võimalikuks lahenduseks alati, kui organisatsioonid teevad strateegilisi või poliitilisi otsuseid, võttes hoolikalt arvesse tehnoloogia eeliseid ja riske.

Mõju ja sihtrühm

Mõju valdkonnad

Majandus

Ettevõtlus

Sihtrühm: ettevõtjad, avalik sektor, teadusasutused

Halduskoormus

Kas lahendusega kaasneb mõju halduskoormusele? Ei

Teatised otseselt halduskoormust kaasa ei too, küll aga nendest tulenevad konkreetsed ELi õigusaktid, mistõttu tuleks hinnata mõju avaliku sektori ja ettevõtete halduskoormusele.

Riigieelarve

Kas lahendusega kaasneb mõju riigieelarvele? Ei

Teatised otseselt riigieelarvelist mõju kaasa ei too, küll aga tuleb hinnata teatistest tulenevate ELi õigusaktide mõju riigieelarvele.

Kaasamine

Kaasata algatustega seotud huvigrupid.