Lithuania - National AI Strategy Update

Overview

The national update to Lithuania’s AI strategy and the parallel plan to establish a National AI Governance Forum were launched by the Ministry of the Economy and Innovation (EIMIN) in late 2024 to align national policy with the European Union’s Artificial Intelligence Act and to stimulate controlled innovation domestically. The update packages strategic objectives (competitiveness, secure public services, language and data resource development) with concrete operational measures: a regulated "AI sandbox" to safely test AI systems, funding for language and SME AI projects, and strengthened state data governance through the State Data Agency’s State Data Lake program. Official public descriptions and progress reports are published by EIMIN and partner agencies (for example the Ministry’s AI policy page and announcements about the DI smėliadėžė). For programmatic details and official announcements see Ministry of the Economy and Innovation — Artificial Intelligence and the DI sandbox page (DI smėliadėžė).

Definitions

For the Strategy update and the Forum the following working definitions apply: "AI system" follows the EU Artificial Intelligence Act definition (software that is developed with machine learning, logic‑based, or statistical approaches and produces outputs such as content, predictions, recommendations or decisions); "high‑risk AI systems" are those categories listed in the EU AI Act and transposed by national implementing measures; "provider" means the natural or legal person that develops an AI system and places it on the market; "user" means the natural or legal person using an AI system under their responsibility; "sandbox" (DI smėliadėžė) means a supervised regulatory testing environment offering restricted, time‑bound testing of AI systems with defined safeguards and reporting obligations. These definitions inform the governance, conformity and market surveillance workstreams under the Forum.

Governance and Institutional Framework

The Governance Forum is conceived as an inter‑ministerial, multistakeholder coordination body chaired by the Ministry of the Economy and Innovation and including the Innovation Agency (sandbox operator), the State Data Agency (data governance and the State Data Lake), the Communications Regulatory Authority (RRT — national contact point and market surveillance authority), the Ministry of Health, the Ministry of Justice, Ministry of Finance and other sectoral bodies. The Forum’s remit includes policy coordination, technical guidance, oversight of national implementation of the EU AI Act, public consultations, and escalation channels for cross‑agency decisions. The Forum will convene working groups (technical, ethics, data access, market surveillance) and publish guidance materials. EIMIN’s public materials set out the principal institutional roles: see the Ministry’s AI page here and the DI sandbox announcement here. The Communications Regulatory Authority’s public site confirms its national supervisory remit and role as contact point for AI market oversight (RRT).

Key Focus Areas

The update prioritises a cluster of policy and operational themes: 1) Governance & oversight — design and mandate of the National AI Governance Forum, cross‑agency workflows, and public consultation mechanisms; 2) Risk management & conformity — adoption of EU high‑risk categories, national channels for conformity assessment and registration, and technical guidance for providers and notified bodies; 3) Safety, testing & evaluation — a regulated AI sandbox (DI smėliadėžė) to test high‑risk and innovative AI systems under supervision with pre‑agreed mitigation measures; 4) Transparency & documentation — obligations on documentation, technical documentation and logs to support explainability and auditing; 5) Data protection & privacy — state data governance measures, secure access procedures, and compliance with national and EU data protection rules to underpin safe public‑sector AI; 6) Market surveillance & enforcement — empowering RRT and designated bodies to perform post‑market monitoring, conformity checks and risk‑based interventions; 7) Cybersecurity & model security — measures and guidance for protecting model integrity, provenance and supply chains; 8) Accountability & liability — processes for incident reporting, redress and coordination with courts and administrative authorities. The Ministry has published related measures and funding to support SME uptake and Lithuanian language resources to lower barriers to entry for domestic AI developers (language projects).

Implementation Framework

Implementation is staged: initial legal amendments and institutional design were submitted by EIMIN (late 2024); parliamentary approval and statutory amendments were progressed into 2025 to permit the sandbox and reassign responsibilities to implementing agencies. The Innovation Agency is mandated to design and run the DI smėliadėžė (sandbox) with participant admission rules, monitoring and reporting templates. The State Data Agency is continuing rollout of the State Data Lake to create secure, auditable channels for public data access in support of vetted sandbox testing and authorised public‑sector AI deployments. RRT will implement market surveillance programs and act as the single national contact point for cross‑border cooperation. The Forum will produce guidance documents, templates for technical documentation and conformity‑assessment roadmaps, and an operational handbook for the sandbox. Official program descriptions and timelines are maintained on EIMIN’s AI pages and the State Data Agency pages (see Sources).

Monitoring and Evaluation

Monitoring is twofold: (1) process monitoring by the Forum to track implementation milestones (legislative transposition, sandbox operational readiness, data lake integrations, stakeholder engagements); and (2) outcome monitoring by sectoral regulators (RRT, health, finance regulators) performing post‑market surveillance to detect conformity deviations, systemic risks and incidents. The Forum will define KPIs (adoption by public bodies, number of sandbox pilots, successful conformity assessments, incidents reported and remediated) and publish periodic status reports. The State Data Agency will provide operational metrics on data maturity and access logs to support evaluation of public‑sector AI deployments.

Penalties, Liability, and Appeals

Enforcement of requirements follows the EU Artificial Intelligence Act’s structure. National implementing measures under review will set administrative procedures for registrations, conformity checks and market surveillance; enforcement powers (orders, corrective measures, penalties) will be exercised by designated national authorities including RRT and sectoral regulators. Providers and deployers of non‑conforming high‑risk systems may face administrative fines, remediation orders and market restrictions in line with EU framework; appeals will be available via administrative and judicial review channels established by Lithuanian administrative law. Official references and preparatory legal measures are published by EIMIN and by the Seimas legislative tracking resources used during the amendment process.

Relationship to Other Instruments

The Strategy update is explicitly designed to implement and align with the EU Artificial Intelligence Act (Regulation (EU) 2024/1689 and related acts), the EU Data Act, the Data Governance Act and the European Interoperability framework. It also links to national digital agendas (National Plan for the Digital Decade) and to public‑sector programs such as the State Data Lake rollout. The DI sandbox is established under national procedural rules to meet the EU AI Act’s sandbox/testing provisions and is coordinated with national data governance instruments to ensure lawful, auditable data access for testing purposes.

International Alignment

Lithuania’s approach seeks close alignment with EU implementation guidance and cross‑border cooperation mechanisms. The Government frames the Forum’s mandate to support interoperability with EU competent authorities, to participate in mutual recognition and exchange of best practices, and to contribute to European capacity building efforts (AI competence centres, research consortia). Initiatives such as the proposed Baltic AI gigafactory and EU‑funded AI centres (LitAI) exemplify regional cooperation to scale research and secure infrastructure. Official announcements by the Ministry describe Lithuania’s intent to integrate national AI infrastructure into broader EU ecosystems.

Implementation Timeline

Sources and References