OECD’s Recommendation of the Council on Artificial Intelligence

(link)

This document contains a non-binding set of standards developed for members of the Organisation for Economic Co-operation and Development. It is primarily foundational and legislative in nature, which explains the language adopted. It includes a set of five principles and five recommendations for trustworthy AI.

RECOGNISING that, at the same time, these transformations may have disparate effects within, and between societies and economies, notably regarding economic shifts, competition, transitions in the labour market, inequalities, and implications for democracy and human rights, privacy and data protection, and digital security;

Principles for responsible stewardship of trustworthy AI

Inclusive growth, sustainable development and well-being

This refers to notions of inclusivity, fairness, environmental well-being and sustainability.

Human-centred values and fairness

These include freedom, dignity and autonomy, privacy and data protection, non-discrimination and equality, diversity, fairness, social justice, and internationally recognised labour rights.

This section also considers safeguards and human oversight.

Transparency and explainability

Transparency here includes the following aspects:

  • Transparency of AI decisions and recommendations
  • General understanding of AI systems
  • Awareness of stakeholders in their interactions with AI
  • Awareness and understanding by those affected by AI-assisted decisions
  • Capacity for human understanding and redress

Robustness, security and safety

This section touches on robustness, safeguards in case of failure, abuse or misuse, traceability throughout the AI and data pipelines, and risk management “to address risks related to AI systems, including privacy, digital security, safety and bias”.

Accountability

The standards hold “AI Actors” accountable for “the proper functioning of AI systems” in accordance with the above principles.

AI actors are those who play an active role in the AI system lifecycle, including organisations and individuals that deploy or operate AI.

Recommendations for national policies and international cooperation

Interestingly, the guidelines explicitly mention “special attention to small and medium-sized enterprises (SMEs)”.

Investing in AI research and development

This section encourages public and private investment in the following areas:

  • General research in trustworthy AI and its technical issues
  • AI-related social, legal and ethical implications and policy issues
  • Open datasets that are representative, free of inappropriate bias and respect privacy and data protection
  • Development of platforms and tools that focus on interoperability and open standards

Note the use of the phrase inappropriate bias.

Fostering a digital ecosystem for AI

To build on the previous recommendation, this section focuses on the development of an ecosystem / infrastructure that facilitates trustworthy AI, sharing of appropriate AI knowledge and “safe, fair, legal and ethical sharing of data”.

Shaping an enabling policy environment for AI

This considers a policy environment that should ideally:

  • Be agile in adopting appropriate legislative measures
  • Facilitate the appropriate and controlled transition of AI technologies from research to deployment
  • Review policy and regulatory frameworks and assessment mechanisms in light of trustworthy AI

Building human capacity and preparing for labour market transformation

This section focuses on the workplace in two main aspects:

  • Preparing for the transformation of the labour market due to displacement by AI
  • Fair use of AI in the workplace to enhance the wellbeing and productivity of workers

International co-operation for trustworthy AI

Finally, this recommendation focuses on international and multidisciplinary cooperation for the following:

  • Appropriate sharing of AI knowledge and expertise
  • Development of global technical, legal and ethical standards
  • Open development of tools, frameworks and datasets
  • Development and sharing of “internationally comparable metrics to measure AI research, development and deployment”