Latvia - AI Development Report (2020)

Overview

The informative report "On the Development of Artificial Intelligence Solutions" (adopted 4 February 2020) is a policy planning document prepared by the Ministry of Environmental Protection and Regional Development (VARAM) and approved by the Cabinet of Ministers. It articulates Latvia's initial national position on AI, explains core AI technologies and methods, surveys international and domestic use cases, and prescribes actions for the next three years to promote AI adoption in public administration and the economy. The report was prepared in response to the EU Coordinated Plan on AI and to support preparation of the national "Digital Transformation Guidelines 2021–2027". The full text is published on the national legislation portal and the Cabinet TAP portal; the official publication page is available at Likumi.lv – Informatīvais ziņojums 'Par mākslīgā intelekta risinājumu attīstību'.

Definitions

The report provides working definitions and taxonomy for AI used in the Latvian public sector context. It distinguishes between narrow (task-specific) AI and general AI, and describes common subfields: machine learning (including deep learning and reinforcement learning), expert systems, natural language processing (NLP), computer vision, robotics and hybrid cyber-physical systems. The document highlights key characteristics: data-driven model training, adaptive behavior, limited explainability for some methods, and the dependence of output quality on training data. The report notes that academic definitions vary and therefore adopts practical operational definitions to guide public sector adoption and risk assessment.

Governance and Institutional Framework

The report designates VARAM as the primary coordinating body responsible for embedding AI priorities into national digital planning and public administration modernization. It tasks ministries and state institutions to evaluate AI opportunities in their sectors and to report use cases, costs and expected benefits. The document instructs VARAM to prepare amendments to the national e-index to include AI indicators and to present recommendations for public communication on AI. Implementation governance is envisioned as a cross-ministerial effort, with the Cabinet of Ministers and the State Chancellery playing oversight roles. The report also explicitly links national activities to EU-level coordination mechanisms and encourages participation in international research and standard-setting bodies (for example the EU High-Level Expert Group outputs and the OECD AI initiatives). Relevant official references appear on the Cabinet's TAP portal and the Likumi.lv publication page (TAP portal – MK protocol).

Key Focus Areas

The report organizes policy actions around several strategic focus areas: (1) Education & research – strengthening AI curricula, upskilling public servants and supporting academic research; (2) Data ecosystems – improving availability of structured, machine-readable public data and repositories to support machine learning; (3) Infrastructure – investing in computing infrastructure including HPC and cloud resources; (4) Public sector pilots – encouraging ministries to trial AI for service automation, analytics and proactive services; (5) Legal & ethical frameworks – preparing guidance on data protection, explainability, and human oversight; (6) Risk management – embedding risk assessment and monitoring for AI deployments; (7) Communication – drafting recommendations to explain AI to the public and build trust; and (8) International cooperation – aligning with EU and OECD instruments. The report emphasizes public sector examples already in use (virtual assistants, number plate recognition systems) and asks ministries to include AI-related costs and benefits in budgeting for 2020–2029.

Implementation Framework

Implementation is presented as phased and collaborative. Immediate tasks include: VARAM to integrate AI into the Digital Transformation Guidelines; ministries to conduct reviews of applicability and prepare internal roadmaps; development of monitoring indicators integrated into the e-index; and targeted funding measures for pilot projects. The report recommends establishing shared services (e.g., a universal virtual assistant for state institutions) and centralized technical support to reduce duplication. It also proposes preparatory work for legal and ethical guidance, including coordination with data protection authorities and cybersecurity agencies to ensure compliance with GDPR and national security requirements.

Monitoring and Evaluation

Monitoring is driven by a set of proposed indicators (AI uptake, number of pilot projects, data availability metrics, skills metrics) to be included in Latvia's e-index. The report calls for periodic reporting to the Cabinet and recommends that implementation be reviewed in connection with the Digital Transformation Guidelines 2021–2027. The State Audit Office (Valsts kontrole) and other oversight bodies are identified as likely evaluators of progress and cost-efficiency in follow-up reviews. The report envisages regular adjustment of actions based on technology evolution and policy feedback.

Penalties, Liability, and Appeals

As an informative, strategic document the report does not itself create sanctioning powers or prescriptive liability rules. It recommends alignment with existing legal frameworks, including the GDPR, civil liability norms and sectoral legislation. The document calls for further work to assess liability and redress mechanisms for AI systems used by the state and recommends that future regulatory measures define accountability and appeals processes where AI affects rights or entitlements.

Relationship to Other Instruments

The report explicitly positions itself to feed into and be operationalized by the "Digital Transformation Guidelines 2021–2027" and aligns with the EU Coordinated Plan on AI, OECD AI policy work and other international standards. It recommends cross-references with national data policy, public procurement rules and sectoral regulations (health, transport, justice) to ensure coherent treatment of AI. The TAP portal record and official publication confirm the report's role as a planning-level instrument rather than a binding law.

International Alignment

International alignment is a core principle: the report references the EU Coordinated Plan (2018) and invites active participation in EU and OECD initiatives. It seeks to ensure that Latvia's approach to AI adoption, ethics and safety is interoperable with EU regulatory developments (including the EU AI regulatory proposals and standards work) and with regional best practices. The report recommends cooperation on data sharing (weather/environment repositories), language technologies and research partnerships.

Implementation Timeline

Sources and References