AI Regulation Hub

Serbia

Serbia has no comprehensive AI Act, but has adopted the Strategy for the Development of AI 2025–2030 (10 January 2025) and Ethical Guidelines for Robust and Accountable AI, supported by the Council for Artificial Intelligence which prepares laws and recommendations.

Key provisions

Strategy for the Development of AI 2025–2030

In force

Adopted by the Government of Serbia on 10 January 2025. Focuses on AI research, development, public-sector adoption, innovation, skills, infrastructure and responsible AI use.

Ethical Guidelines for Robust and Accountable AI

In force

Define robust and accountable AI as technically robust and safe, legally compliant and ethically aligned. Cover human agency, technical reliability, privacy and data governance, transparency, non-discrimination, social/environmental well-being and accountability.

AI assessment questionnaire

In force

Practical questionnaire that organisations can use to assess AI systems during planning, beta testing and later lifecycle stages — does not replace legal compliance.

Council for Artificial Intelligence

In force

Monitors AI development and application, prepares proposals, recommendations and standards, and organises and monitors preparation of laws and regulations concerning AI.

Detailed overview

Serbia does not currently have a comprehensive AI Act, but it has a developed AI policy framework and official ethical guidelines. The Government of Serbia adopted the Strategy for the Development of Artificial Intelligence for the period 2025–2030 on 10 January 2025.

Serbia's AI strategy builds on its earlier AI strategy and positions the country as a regional AI leader. The strategy focuses on AI research, development, public-sector adoption, innovation, skills, infrastructure and responsible AI use.

Serbia has also adopted Ethical Guidelines for the Development, Implementation and Use of Robust and Accountable Artificial Intelligence. The guidelines define robust and accountable AI as AI that is technically robust and safe, compliant with law, and compliant with ethical principles and values.

The Serbian guidelines identify practical requirements for AI governance. These include human agency and control, technical reliability and security, privacy and data governance, transparency, diversity, non-discrimination and equality, social and environmental well-being, and accountability. They also include a questionnaire that organisations can use to assess AI systems during planning, beta testing and later lifecycle stages.

The questionnaire does not replace legal compliance. The guidelines state that the questionnaire is not a guide to the Serbian legal system and that completing it does not relieve organisations of legal obligations and responsibilities.

Serbia has also established a Council for Artificial Intelligence. The Council monitors AI development and application, prepares proposals, recommendations and standards, and organises and monitors preparation of laws and regulations concerning AI development and use.

Serbia does not currently have one AI-specific penalty table. Penalties depend on the underlying law breached, such as personal-data protection, cybersecurity, consumer protection, financial regulation, healthcare regulation, employment law, intellectual-property law, civil liability or criminal law.

Practical requirements & details

Sourced from the Strategy for the Development of Artificial Intelligence 2025–2030 (adopted 10 January 2025), the Ethical Guidelines for the Development, Implementation and Use of Robust and Accountable AI, and materials of the Council for Artificial Intelligence.

Strategy focus

  • AI research, development, public-sector adoption, innovation, skills, infrastructure and responsible AI use.
  • Positions Serbia as a regional AI leader.

Robust and accountable AI

  • AI must be technically robust and safe.
  • AI must be compliant with law.
  • AI must be compliant with ethical principles and values.

Practical governance requirements

  • Human agency and control.
  • Technical reliability and security.
  • Privacy and data governance.
  • Transparency.
  • Diversity, non-discrimination and equality.
  • Social and environmental well-being.
  • Accountability.

AI assessment questionnaire

  • Use to assess AI systems during planning, beta testing and later lifecycle stages.
  • Does not replace legal compliance — completing it does not relieve organisations of legal obligations.

Council for Artificial Intelligence

  • Monitors AI development and application.
  • Prepares proposals, recommendations and standards.
  • Organises and monitors preparation of laws and regulations concerning AI.

Penalties

  • No AI-specific penalty table.
  • Penalties depend on the underlying law: data protection, cybersecurity, consumer, financial, healthcare, employment, IP, civil or criminal law.

Ready to launch legally?

Book a 30-minute consultation. We'll map your licensing path and tell you exactly what's required, in plain language.