Austria - AI Measures Package

Overview

The KI-Maßnahmenpaket is a national policy package presented on 21 September 2023 that combines immediate administrative measures, statutory amendments and capacity-building initiatives to prepare Austria for the EU AI Act and to foster trustworthy AI deployment. Its headline measures were the announcement of a national KI-Servicestelle to be hosted at the Rundfunk und Telekom Regulierungs-GmbH (RTR), the introduction of a mandatory AI labelling/notification approach for public transparency, and a set of competence-building initiatives for citizens, businesses and schools. The presentation and rationale are documented in the Austrian Federal Ministry of Finance press release (see BMF press release: Tursky präsentiert KI-Maßnahmenpaket) and were followed by legislative amendments creating the legal basis for the service desk (BGBl. I Nr. 6/2024). The KI-Servicestelle has since been implemented at the RTR as the operational contact point for the public and stakeholders (RTR KI-Servicestelle).

Definitions

For the purposes of the measures package and the implementing statutory amendments, key terms are defined in line with EU-level practice: "KI" (AI) refers broadly to software systems that, by means such as machine learning or pattern recognition, produce outputs such as content, predictions, recommendations or decisions affecting environments or people. "Provider" denotes the natural or legal person that develops or places an AI system on the market; "operator" or "user" denotes the entity that deploys an AI system in Austria. "High-risk" follows a risk-based approach consistent with the EU AI Act's categories (health, safety-critical systems, HR selection, biometric ID, law enforcement functions etc.). The statutory amendments reference these definitions while the RTR guidance supplies practical interpretive material for Austrian actors.

Governance and Institutional Framework

The package establishes a two-step national governance model: an immediately deployable KI-Servicestelle hosted within the RTR as an accessible advisory and coordination hub, and a planned transition to a dedicated national AI authority with regulatory, certification and market surveillance competencies once that role becomes necessary under the EU AI Act. The legal footing for the KI-Servicestelle was created by amending the KommAustria Act and the Telecommunications Act 2021; these changes are codified in BGBl. I Nr. 6/2024 and are described on the RIS portal (BGBl. I Nr. 6/2024). The RTR pages provide operational detail, including responsibilities, contact points and the creation of an AI Advisory Board to provide expert counsel (RTR KI-Servicestelle). The governance design emphasizes service-orientation and coordination across the Federal Government, sectoral ministries and EU-level bodies to ensure alignment with the EU AI Act and other EU data rules.

Key Focus Areas

The package concentrates on transparency, competence-building, preparatory compliance support, and institutional readiness. First, transparency: the Government announced a national AI labelling/"Kennzeichnungspflicht" to ensure citizens know when they interact with AI and to create baseline disclosure practices. Second, competence: the measures finance or coordinate education and public-awareness campaigns, school pilots, and training for public servants and SMEs to build practical know-how. Third, preparatory compliance: the KI-Servicestelle publishes guidance, runs events, operates an AI Act chatbot and creates practical materials to assist providers and deployers in conducting risk assessments, technical documentation and conformity steps. Fourth, institutional readiness: statutory amendments create the RTR-based Servicestelle's legal basis and foreseen escalation path to a national AI authority responsible for market surveillance, conformity assessment and enforcement once the EU framework is applicable. Sectoral emphasis is placed on media (due to RTR’s remit), telecoms, healthcare, energy and public administration—areas where AI intersects with fundamental rights, safety or critical infrastructure. The package also signals close coordination with EU-level rules such as the Data Governance Act and the Data Act to ensure data availability and lawful use in AI pipelines (BMF press release).

Implementation Framework

Implementation is intentionally phased: immediate administrative actions (information hub, advisory board, labeling pilot concepts and trainings) were launched in late 2023/early 2024; legal amendments providing a statutory basis for the Servicestelle were enacted and published as BGBl. I Nr. 6/2024 on 26 February 2024; subsequent steps include development of enforcement and market surveillance capacity and the formal creation of a national AI authority when required by the timing of the EU AI Act. Implementation relies on cross-ministerial coordination, use of RTR's operational expertise in telecommunications and media, and on public consultations. The RTR offers tools and materials (FAQs, chatbot, guidance documents) and convenes stakeholder events to translate EU obligations into Austrian practice. Funding and resource allocation combine national budget lines and the existing RTR institutional funding model, supplemented by targeted governmental programs for education and SME support.

Monitoring and Evaluation

Monitoring uses a mixed-methods approach: the RTR Servicestelle collects uptake metrics (events, downloads, inquiries), produces sectoral reports (e.g., HR survey, sectoral workshops) and publishes evidence-based guidance; ministries will track policy KPIs (competence training reach, labelling adoption, number of advisory interventions). The national AI Advisory Board provides qualitative advice and periodic reviews of the measures package. In addition, Austria anticipates relying on EU-level implementation milestones and conformity outcomes under the AI Act to evaluate whether and when the Servicestelle should be transitioned into a fully empowered national AI authority with market surveillance powers. Regular public reporting and stakeholder consultations are foreseen to ensure transparency of progress.

Penalties, Liability, and Appeals

The KI-Maßnahmenpaket itself is primarily a policy and preparatory legislative package; it does not, on its own, set exhaustive new penalty regimes beyond existing legal powers. The package prepares Austria to enforce the EU AI Act’s sanctioning and liability mechanisms once the EU framework becomes applicable. The national amendments enable future market surveillance and administrative actions via RTR and related authorities. Liability for harms caused by AI systems remains a mix of existing civil liability regimes and forthcoming EU-level rules; administrative penalties for breaches of mandatory transparency or conformity obligations will follow the EU-tiered approach (significant fines for severe breaches) and national implementing instruments. The RTR guidance clarifies that enforcement will be risk-based and proportionate while ensuring remedial pathways and appeal rights consistent with Austrian administrative law.

Relationship to Other Instruments

The package explicitly situates itself in the EU regulatory landscape: it is designed to complement and operationalize obligations under the EU AI Act, the Data Governance Act and the Data Act. National statutory changes to KommAustria and the Telecommunications Act were deliberate steps to create a legally anchored Servicestelle that can support AI Act implementation. The measures also interact with sectoral laws (healthcare regulation, labour law, media law) where AI use implicates specific rights or safety regimes. The package aims to avoid duplication by offering guidance and capacity building rather than setting separate, conflicting substantive AI rules that would diverge from harmonized EU law.

International Alignment

Austria’s KI-Maßnahmenpaket is explicitly aligned with EU developments and positions Austria as an early implementer of preparatory national infrastructure—alongside Spain and the Netherlands—by establishing an AI Service Desk and advisory structures. The package also emphasises engagement with EU institutions and international initiatives to ensure compatibility with EU AI Act timelines and standards and to support the interoperability of conformity assessments and market surveillance across borders. Austria’s approach balances promoting competitiveness with rights protection, reflecting EU policy priorities and fostering international cooperation in research, standardization and enforcement practices.

Implementation Timeline

Sources and References