Hungary AI Strategy 2025

Overview

Magyarország Mesterséges Intelligencia Stratégiája (2025-2030) is Hungary’s renewed national AI policy framework. Dated 2025-09-03, it revises the country’s 2020 AI strategy in response to rapid technological developments, particularly generative AI, expanding European digital regulation, changing labour-market conditions, cybersecurity risks, and the strategic importance of data sovereignty. It is a strategy rather than a primary law, decree, or directly enforceable regulatory instrument. Its central purpose is to prepare Hungarian society, public institutions, researchers, and businesses to use AI productively while addressing security, ethical, legal, and social risks.

The strategy frames AI as a horizontal technology and an economic-development tool. It retains a 2030 horizon and sets high-level objectives of responsible use, national and international cooperation, increased competitiveness, and practical public benefit. Its principal policy architecture consists of six foundational pillars: regulation and security; infrastructure; education and skills; the data economy; research, development and innovation; and the encouragement and development of applications. It also identifies three priority domains: “AI for society,” “AI for technology,” and “AI for business.”

Definitions

The document uses “MI,” short for mesterséges intelligencia, to mean artificial intelligence. It describes AI as extending beyond algorithmic systems trained and improved on data, increasingly simulating aspects of human understanding, learning, and problem-solving. The strategy distinguishes AI systems from general-purpose AI models and recognises large language models and generative AI as important developments requiring specific attention to transparency, user protection, culture, language, cybersecurity, and misinformation risks.

Several institutional and operational concepts are central. The proposed MI Hivatal, or Hungarian Artificial Intelligence Office, is intended to perform AI Act-related national functions. The proposed MI Tanács, or Hungarian Artificial Intelligence Council, is intended to provide opinions, recommendations, and guidance. A regulatory sandbox is envisaged as a controlled setting for testing AI systems and compliance. The document also uses “adattárca” to describe a data-wallet concept designed to let citizens manage data-related permissions and sharing choices within the GDPR framework.

Governance and Institutional Framework

The strategy calls for a coordinated national governance model centred on implementation of Regulation (EU) 2024/1689, the EU AI Act. It states that the AI Act is directly applicable and should not be duplicated through a parallel Hungarian AI regime. Instead, Hungary should establish national implementation structures, sectoral arrangements, supervisory mechanisms, and practical guidance. The planned AI Office is intended to perform market-surveillance and notifying-authority functions, act as the single point of contact towards EU institutions, operate a regulatory sandbox, and support the work of the AI Council.

The AI Council is conceived as a multi-participant body supporting national policy, monitoring and coordinating AI Act implementation activities, issuing recommendations compatible with EU guidance, and contributing to reports on implementation and AI use. The strategy also expects close cooperation between AI governance bodies and cybersecurity authorities and incident-response centres, especially regarding high-risk AI systems, critical infrastructure, data-security requirements, information exchange, and coordinated supervision. The Hungarian Academy of Sciences is expected to be involved through a scientific advisory body.

Key Focus Areas

“AI for society” concerns public services, health, education, transport, energy, public administration, law enforcement, civil protection, and defence-related capacities. The strategy supports Hungarian-language automated administration, user-feedback mechanisms in AI-based public services, AI-supported health research and decision support, energy forecasting, adaptive traffic management, and inclusive AI-supported learning. It stresses that public-sector AI should be transparent, subject to human oversight, and accompanied by complaint or appeal possibilities where individuals object to AI involvement.

“AI for technology” focuses on structured and accessible data, foundation models, machine perception, robotics, test environments, autonomous systems, secure infrastructure, and open-source development. “AI for business” focuses on productivity, smart manufacturing, predictive maintenance, supply chains, finance, agriculture, and SME adoption. The strategy specifically seeks to make AI tools, advisory services, training, proof-of-concept financing, standardised integration tools, and Hungarian-language models more accessible to enterprises, especially smaller businesses that currently face adoption, cost, skills, and compliance barriers.

Implementation Framework

The implementation model is designed to be agile, modular, iterative, and feedback-based. The strategy assigns or envisages roles for ministries, the AI Commissioner, the Ministry for National Economy, the National Data Asset Agency, the Artificial Intelligence National Laboratory, the Digital Government Development and Project Management entity, higher-education institutions, research bodies, professional organisations, and the AI Coalition. The AI Coalition is described as a broad ecosystem forum bringing together government, business, research, education, and social stakeholders.

Infrastructure implementation includes expansion of high-performance computing, integrated hybrid cloud capacity, data access protocols, AI development tools, protected sandboxes, testbeds, marketplaces, and automated access-control systems. The strategy calls for a Hungarian AI Factory within the EuroHPC context and seeks rapid access to AI-optimised resources. It also anticipates data-quality, anonymisation, access-control, monitoring, incident-management, and security-by-design measures. Businesses and public entities are expected to treat cybersecurity, supply-chain integrity, model vulnerabilities, adversarial attacks, and resilient operations as integral elements of AI development and deployment.

Monitoring and Evaluation

The strategy sets 2030 headline indicators: a 15% AI-induced GDP increase relative to 2020, a 26% average productivity increase in the Hungarian corporate sector relative to 2020, and one million citizens performing higher-value AI-supported work following job or role changes. Additional targets include AI-supported education for 2.5 million citizens, AI use by 40% of SMEs, a 40-petaflop Hungarian supercomputer, at least 200 AI start-ups, 1,000 productised AI applications, and 200 proof-of-concept projects.

Annex 1 establishes operational follow-up actions. The AI Commissioner is responsible for developing the methodology for the strategy’s indicator system by 2025-11-30, tracking technological, economic, social, and legal changes annually, and monitoring indicator performance annually. The AI Coalition and relevant organisations are to conduct necessary measurements annually. The strategy’s executive summary says it will be renewed annually, while Annex 1 provides that renewal should occur as necessary but at least every two years. This creates a policy expectation of continuing review and recalibration rather than a fixed one-time implementation model.

Penalties, Liability, and Appeals

The strategy does not itself establish criminal offences, administrative fines, civil damages rules, or a standalone appeal procedure. As a policy instrument, it proposes institutional, legal, ethical, and technical measures rather than directly imposing sanctions on AI providers, deployers, or users. Binding duties, enforcement powers, and penalties relevant to many AI systems arise primarily from the EU AI Act, GDPR, cybersecurity law, sector-specific requirements, and other applicable Hungarian and EU instruments.

Nevertheless, the strategy treats accountability as a substantive implementation priority. It supports clarifying legal and ethical responsibility among developers, providers, and users, especially for autonomous decision-making systems. It proposes documentation and disclosure practices for high-risk systems, independent audits, standardised testing, monitoring for algorithmic bias, and human-in-the-loop safeguards for critical decisions. For public-sector systems, it calls for clear information about the extent of AI involvement and for complaint or appeal opportunities where citizens challenge AI-supported decisions or services.

Relationship to Other Instruments

The strategy is expressly designed to operate alongside EU law rather than replace it. Its principal regulatory reference is the EU AI Act, Regulation (EU) 2024/1689. It also refers to the GDPR, the Data Act, the Data Governance Act, the Open Data framework, the European Health Data Space regulation, and EU cybersecurity requirements. It identifies national implementation and operationalisation of these rules as priorities, including data governance, secure public-data reuse, sector-specific frameworks, high-risk AI oversight, and data-sharing infrastructure.

Within Hungary, the strategy builds on the 2020 national AI strategy adopted by Government Decision 1573/2020 (IX. 9.). It refers to Government Decision 1301/2024 (IX. 30.), concerning national AI Act-related functions, and Government Decision 1149/2025 (V. 14.), under which market-surveillance responsibilities are to be performed by the Ministry for National Economy and notifying-authority responsibilities by the National Accreditation Authority. It also identifies interaction with Hungary’s cybersecurity strategy and health-data reforms effective from 2026-01-01.

International Alignment

The strategy presents Hungary as a participant in European and global AI cooperation while preserving digital and data sovereignty. It supports the EU digital single market, active participation in EU AI governance, and engagement with European research and infrastructure initiatives. Hungary is expected to participate in the EU AI Board’s work and pursue involvement in EuroHPC AI Factory infrastructure. The document also calls for collaboration in international research networks, cross-border data ecosystems, and bilateral investment and innovation partnerships.

Its ethical and policy orientation refers to OECD AI recommendations, UNESCO’s AI ethics recommendation, the G7 Hiroshima AI Process, European Commission work, and Council of Europe AI-related work. The strategy recognises the need to participate in international AI value chains while protecting valuable Hungarian data and managing cross-border risks. It therefore combines support for international cloud, AI, research, and commercial partnerships with an emphasis on lawful data flows, transparent governance, strong cybersecurity, and the protection of Hungarian language and cultural context in AI systems.

Implementation Timeline

Compliance Checklist

Sources and References