Detailed overview
Latvia is regulated by the EU AI Act and has also created a national AI institutional framework through the Artificial Intelligence Centre. The EU AI Act applies directly in Latvia and governs prohibited AI, high-risk AI, transparency obligations and general-purpose AI models.
A high-risk AI system is not banned, but it must satisfy strict legal controls. These include risk management, data governance, documentation, transparency, human oversight, accuracy, robustness, cybersecurity, conformity assessment and post-market monitoring. High-risk AI may include AI used in employment, education, credit, insurance, biometric identification, public services, critical infrastructure, migration, law enforcement and justice.
Latvia's Ministry of Economics announced that the Artificial Intelligence Center would begin operations by 31 March 2025. The centre is intended to become a leading platform for the development and implementation of AI solutions and to bring together scientists, academia, the public sector and businesses.
Latvia's AI Centre is also connected with AI sandbox development. Official materials state that the Artificial Intelligence Development Law would simplify the creation of a special regulatory environment, or sandbox, for AI. A sandbox allows companies to test and develop AI solutions in a safe and controlled setting before broader market deployment.
AI systems in Latvia may also be regulated under GDPR, cybersecurity law, employment rules, consumer law, financial regulation, healthcare regulation, public-sector procurement rules and sector-specific obligations. AI systems involving personal data, Latvian-language models, public-sector use, biometric systems, healthcare, finance or employment should be reviewed under the relevant legal regime.
Penalties for EU AI Act breaches follow the EU AI Act. Prohibited AI breaches may be fined up to EUR 35 million or 7% of worldwide annual turnover. Other AI Act breaches may be fined up to EUR 15 million or 3%, and misleading information to authorities may be fined up to EUR 7.5 million or 1%.
Practical requirements & details
Sourced from Regulation (EU) 2024/1689 (the EU AI Act) and Latvia's Artificial Intelligence Development Law + Artificial Intelligence Centre institutional framework.
EU AI Act categories
- Prohibited AI — banned.
- High-risk AI — strict legal controls (not banned).
- Transparency obligations — disclosure for chatbots/deepfakes.
- GPAI models — documentation, transparency and copyright-policy duties.
High-risk areas
- Employment, education, credit, insurance, biometric identification, public services, critical infrastructure, migration, law enforcement and justice.
Compliance code for high-risk AI
- Risk management, data governance, documentation, transparency, human oversight, accuracy, robustness, cybersecurity, conformity assessment and post-market monitoring.
Artificial Intelligence Centre
- Ministry of Economics announced operations by 31 March 2025.
- Leading platform for AI development and implementation.
- Brings together scientists, academia, the public sector and businesses.
AI Development Law — sandbox
- Special regulatory environment for AI testing — controlled setting before broader market deployment.
Overlapping legal regimes
- GDPR, cybersecurity law, employment rules, consumer law, financial regulation, healthcare regulation, public-sector procurement rules and sector-specific obligations.
Penalties
- EUR 35 million or 7% of worldwide annual turnover — breaches of prohibited AI rules.
- EUR 15 million or 3% of worldwide annual turnover — breaches of many other AI Act operator obligations.
- EUR 7.5 million or 1% of worldwide annual turnover — supplying incorrect, incomplete or misleading information to authorities.
Related entries
See also the European Union entry, which covers the EU AI Act (Regulation (EU) 2024/1689) — the substantive framework that this jurisdiction implements and supervises domestically.