Skip to main content
Please wait...

 Database index

Soft law

OECD AI Principles

Country
Global
Full name of the document
Organization for Economic Co-operation and Development (OECD) - Recommendation of the Council on Artificial Intelligence
Body / organism enacting the document
OECD Council
Nature of the body / organism
International body
Type of document
Ethical principles
Date
Project area
AI and consumer markets
AI and data protection
Cross sector, general scope
Law area
Privacy / data protection
Consumer protection
Fundamental rights protection
Product safety

Summary of the document

The principles are designed to promote innovation and trust in AI while ensuring respect for human rights and democratic values. They are structured around five foundational pillars:
1.    Inclusive Growth, Sustainable Development, and Well-being: AI should be designed in a way that fosters economic growth, promotes social inclusion, and ensures sustainability. This involves creating AI systems that can contribute positively to societal goals and help address global challenges.

2.    Human-centered Values and Fairness: The development, deployment, and use of AI should respect human rights, diversity, and democratic values. This includes ensuring that AI systems do not discriminate and that they work to promote fairness and social justice.

3.    Transparency and Explainability: AI systems should be transparent, and their operations should be understandable by the people who use them or are affected by them. This principle emphasises the importance of clarity about AI decision-making processes and the need for systems that can be audited and scrutinised.

4.    Robustness, Security, and Safety: AI systems must be secure, reliable, and robust enough to handle errors or inconsistencies during all phases of their life cycle. This requires that AI systems are designed with safety in mind, including safeguards to prevent unauthorised access and ensuring they are resilient to attacks and errors.

5.    Accountability: Organisations and individuals developing, deploying, or operating AI systems should be held accountable for their functioning in line with the above principles. This includes mechanisms for responsibility and redress if AI systems cause harm or do not comply with regulatory and ethical standards.

Type of addressees
Unspecified addressees
Territorial scope
Global
Situations involved

The OECD AI Principles, adopted in May 2019, represent a significant international initiative to guide the responsible stewardship of trustworthy Artificial Intelligence

AI system(s) involved
  • All types of AI systems
  • General purpose AI
  • Specific intended purpose AI
  • Generative AI
  • Machine learning
  • Deep learning
Fundamental rights involved
  • Right to data protection
  • Right to privacy
  • Right to private and family life
  • Right to non-discrimination
  • Sustainability
Principles expressly addressed
  • Accountability
  • Equality
  • Explainability
  • Non-discrimination
  • Rule of law
  • Transparency
Connection with hard law

Convention on the Organisation for Economic Co-operation and Development of 14 December 1960
 

Possible legal force and impact on national/supranational legal system

The OECD Recommendation on AI is not legally binding on member countries. It functions as a set of recommended best practices with the following potential areas of impact:

1.    Influence on national legislation through its recognition as a global standard for ethical AI development.
2.    Providing a framework for policymakers to consider when creating national laws and regulations around AI.
3.    Due to its widespread adoption by OECD members, the recommendation can encourage international alignment on AI governance principles

Case author
Ivo Emanuilov
Researcher
LIBRe Foundation