Skip to main content

    AI by country · Northern Europe

    AI in Latvia: the rules, the readiness and where to start

    Last checked: 29 October 2026

    What should a company in Latvia know before starting with AI?

    In Latvia, AI that uses personal data falls under the GDPR and the Personal Data Processing Law of 2018, supervised by the Data State Inspectorate. The EU AI Act applies directly, and a 2025 government plan gives that Inspectorate the main role for banned and high-risk AI. Latvia also set up an Artificial Intelligence Centre by law in March 2025, with a task to run a regulatory testing environment for AI.

    At a glance

    Latvia at a glance

    Key facts about Latvia
    Data protection lawGDPR plus the Personal Data Processing Law (Fizisko personu datu apstrādes likums), adopted 21 June 2018
    RegulatorDatu valsts inspekcija (Data State Inspectorate); the Consumer Rights Protection Centre (PTAC) is listed as AI Act contact point
    AI policyEU AI Act (Regulation (EU) 2024/1689), phased; Artificial Intelligence Centre Law, adopted 6 March 2025
    Time zoneEET, UTC+2 in winter and EEST, UTC+3 in summer (India is 3.5 or 2.5 hours ahead)
    CurrencyEUR (euro)
    Business languageLatvian, with English and Russian also used in business

    Why now

    Why Latvia is worth watching

    Each point links to the document it comes from.
    • Eurostat reports that 12.21% of Latvian enterprises with 10 or more staff used at least one AI technology in 2025, up from 8.83% in 2024. The EU figure was 19.95%, so most Latvian firms are still at the start.[8]

    • The Artificial Intelligence Centre Law, adopted on 6 March 2025, sets up a foundation to bring together the public sector, companies and universities, and gives it the task of providing a special regulatory environment for developing AI systems.[3]

    • On 25 February 2025 the Cabinet examined a plan for the AI Act that gives the Data State Inspectorate market surveillance of banned practices and high-risk AI, and the Bank of Latvia the AI used by financial institutions it supervises.[4]

    • The European Commission lists the Consumer Rights Protection Centre (PTAC) as Latvia's single point of contact for AI Act market surveillance.[6]

    • The EU's Digital Omnibus on AI entered into force on 27 July 2026 and moved high-risk duties for Annex III uses to 2 December 2027.[5]

    The rules

    The rules that matter for AI

    What each law or policy means when you build or buy an AI system.

    General Data Protection Regulation (EU) 2016/679

    In force2016People have a right not to be subject to a decision based solely on automated processing that has legal or similarly significant effects, with listed exceptions (Article 22). High-risk processing needs a data protection impact assessment (Article 35), breaches must be reported within 72 hours where feasible (Article 33), and fines can reach EUR 20 million or 4% of worldwide turnover (Article 83).[1]

    Personal Data Processing Law (Fizisko personu datu apstrādes likums)

    In force2018Adopted on 21 June 2018 and published on 4 July 2018, it took effect the day after proclamation. It fills in the GDPR for Latvia, for example setting 13 as the age at which a child can consent to online services (section 33). We read the original text; later amendments were not checked.[2]

    Artificial Intelligence Centre Law (Mākslīgā intelekta centra likums)

    In force2025Adopted on 6 March 2025 and published on 19 March 2025. It creates the Artificial Intelligence Centre as a foundation under VARAM oversight. Its tasks include guidance for the public and private sectors, AI skills, security risks, opinions on deepfakes in election periods, and a special regulatory environment for AI development.[3]

    EU Artificial Intelligence Act, Regulation (EU) 2024/1689, as amended by Regulation (EU) 2026/1744

    Phased2024Bans and AI literacy apply from 2 February 2025, general-purpose AI rules from 2 August 2025, and most other rules from 2 August 2026. High-risk duties now apply from 2 December 2027 (Annex III) and 2 August 2028 (AI in regulated products). Latvia's split of supervisory roles follows a 2025 Cabinet plan; we could not confirm an adopted implementing law as of 29 October 2026.[7]

    Where AI fits first

    Where AI fits first

    Banking and finance

    Document reading, transaction monitoring and customer service in Latvian, English and Russian.

    Under the 2025 Cabinet plan, the Bank of Latvia would supervise AI used by the financial institutions it oversees. Credit scoring is an Annex III high-risk use from 2 December 2027.[4]

    Public administration

    Drafting answers to citizens, searching legal texts and summarising documents.

    The Artificial Intelligence Centre's tasks include guidance for the public sector and projects in state administration.[3]

    Logistics and transit

    Reading customs and shipping documents and forecasting volumes.

    We found no Latvian AI rule specific to this sector. The GDPR applies where staff or customer data is used.[1]

    Education

    Lesson materials and feedback on student work, with teachers deciding grades.

    The Cabinet plan names the State Service for Quality of Education among market surveillance bodies. AI used for admission or grading is high risk from 2 December 2027.[4]

    Your first steps

    Your first steps

    1. Pick one or two narrow AI uses to start with, and note whether either touches hiring, credit, education or biometrics.
    2. Train the staff who use AI tools, since the AI literacy duty has applied since 2 February 2025.
    3. Run a data protection impact assessment before AI processes personal data at scale, and keep a person reviewing significant decisions.
    4. Prepare to report personal data breaches to the Data State Inspectorate within 72 hours.
    5. Label chatbots and AI-generated content where the AI Act transparency rules apply from 2 August 2026.
    6. Watch the Artificial Intelligence Centre for its guidance and regulatory testing environment.
    7. Use a GDPR transfer mechanism before personal data goes to a country without an adequacy decision.

    AI services for Latvia

    AI consulting and development for teams in Latvia

    We work remotely with companies in Latvia, in your time zone, and keep your data under the rules described above.

    AI agents and automation

    Agents that finish multi-step work across your tools, with people approving what matters.

    Document Q&A (RAG)

    Answers from your own documents, with the source shown for every answer.

    AI governance and data protection

    Policies, review steps and audit trails that fit your local data law.

    Software and cloud engineering

    Web apps, integrations and cloud platforms built around your AI use case.

    How we work with you

    How we work with you

    We work remotely with clients worldwide, in the client's time zone. India (UTC+5:30) is 3.5 hours ahead of Latvia in winter (UTC+2) and 2.5 hours ahead in summer (UTC+3). Taking a 9:00 to 17:00 day in Latvia and 9:30 to 18:30 in India, the overlap is about 6 hours in winter (09:00 to 15:00 Latvian time) and 7 hours in summer (09:00 to 16:00). India has no EU adequacy decision, so any personal data we handle needs a GDPR transfer mechanism such as standard contractual clauses.

    Questions

    Common questions about AI in Latvia

    Who will enforce the AI Act in Latvia?

    Under the plan the Cabinet examined in February 2025, the Data State Inspectorate covers banned practices and high-risk AI, the Bank of Latvia covers supervised financial institutions, and other bodies cover their own sectors. The Commission lists PTAC as the single point of contact. We could not confirm the final implementing law.

    What is the Artificial Intelligence Centre?

    A foundation set up under a law adopted on 6 March 2025. It brings together the state, companies and universities, gives guidance, works on AI skills and is tasked with a special regulatory environment for AI development.

    Does the AI Act apply to small Latvian firms?

    Yes, if they provide or use AI systems in scope. Small firms get lower fine caps under Article 99, and most companies that only use AI tools will deal mainly with AI literacy and transparency duties.

    When do high-risk rules start?

    After the Digital Omnibus on AI, high-risk duties for Annex III uses such as hiring, credit and education apply from 2 December 2027. For AI in products covered by EU product law, they apply from 2 August 2028.

    Complimentary tools

    Try these first

    Three complimentary tools from our AI Toolkit that suit a team starting out.

    AI readiness score

    Answer a few questions and see where your team stands, in about five minutes.

    AI use case triage

    Test an idea against data, risk and effort before you spend on it.

    AI project cost estimator

    Get a rough build and run cost range for a first AI project.

    Read next

    Sources

    Sources

    Facts checked on 29 October 2026. Numbers in square brackets above match this list.
    1. [1] Publications Office of the European Union, 2016-04-27. Regulation (EU) 2016/679 (General Data Protection Regulation), Official…(publications.europa.eu)
      Full title and what we used it for

      Regulation (EU) 2016/679 (General Data Protection Regulation), Official Journal L 119, 4.5.2016. Primary. Official Journal text. Used for Articles 22, 33, 35 and 83.

    2. [2] Latvijas Vēstnesis (official gazette), 2018-07-04. Fizisko personu datu apstrādes likums, 21 June 2018(vestnesis.lv)
      Full title and what we used it for

      Fizisko personu datu apstrādes likums, 21 June 2018. Primary. Original published text. Used for adoption and publication dates, entry into force and section 33. Likumi.lv, which holds the consolidated text, could not be reached from our location, so later amendments were not checked.

    3. [3] Latvijas Vēstnesis (official gazette), 2025-03-19. Mākslīgā intelekta centra likums, 6 March 2025(vestnesis.lv)
      Full title and what we used it for

      Mākslīgā intelekta centra likums, 6 March 2025. Primary. Used for sections 1 to 4 (purpose, legal status, tasks) and the adoption and publication dates. Our reading of the Latvian text is our own translation.

    4. [4] Ministry of Smart Administration and Regional Development (VARAM), 2025-02-25. Latvia takes step towards safe and responsible artificial intelligence:…(varam.gov.lv)
      Full title and what we used it for

      Latvia takes step towards safe and responsible artificial intelligence: government examines implementation plan for AI Act. Primary government release. Used for the planned split of supervisory roles. It describes a plan, not an adopted law.

    5. [5] European Commission, 2026-10-19. AI Act: regulatory framework for AI(digital-strategy.ec.europa.eu)
      Full title and what we used it for

      AI Act: regulatory framework for AI. Primary. Used for the timeline and the Digital Omnibus on AI (Regulation (EU) 2026/1744 of 8 July 2026) entering into force on 27 July 2026.

    6. [6] European Commission, 2026-09-07. Market Surveillance Authorities under the AI Act(digital-strategy.ec.europa.eu)
      Full title and what we used it for

      Market Surveillance Authorities under the AI Act. Primary. A list of single points of contact for market surveillance. Used for PTAC.

    7. [7] European Commission, AI Act Service Desk, 2026-10-29. AI Act, Article 113 (Entry into force and application)(ai-act-service-desk.ec.europa.eu)
      Full title and what we used it for

      AI Act, Article 113 (Entry into force and application). Primary text as consolidated after the Digital Omnibus on AI. Article 99 (https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-99) was used for fines and the SME rule, and Annex III (https://ai-act-service-desk.ec.europa.eu/en/ai-act/annex-3) for high-risk areas.

    8. [8] Eurostat, 2026-06-15. Artificial intelligence by size class of enterprise (isoc_eb_ai)(ec.europa.eu)
      Full title and what we used it for

      Artificial intelligence by size class of enterprise (isoc_eb_ai). Primary statistics. Enterprises with 10 or more staff, excluding the financial sector, using at least one AI technology.

    9. [9] European Commission, 2026-10-29. Adequacy decisions(commission.europa.eu)
      Full title and what we used it for

      Adequacy decisions. Primary. Used to confirm that India has no EU adequacy decision.

    This is general information, not legal advice. We could not check Latvia's consolidated laws or confirm a final AI Act implementing law, and the AI Act dates changed in 2026. Check with a qualified Latvian lawyer before you act. Facts checked on 29 October 2026.

    Starting with AI in Latvia?

    Talk to an engineer. The first call is free, and we start with a readiness assessment before any budget is committed.