One of the most trusted global institutions on intergovernmental regulations and standards is the OECD (Organisation for Economic Co-operation and Development). In May 2019, the OECD set a global precedent by adopting the first intergovernmental standard on AI. This initiative wasn't in isolation; it was a response to the mounting need for a harmonized approach towards the burgeoning field of AI. The principles were designed to foster innovation in AI while ensuring it's trustworthy, respecting human rights, and adhering to democratic values.
In the swiftly evolving landscape of Artificial Intelligence (AI), understanding and implementing international standards is crucial for any company aiming to be a responsible AI actor. Here, we delve into the OECD’s AI recommendations and explore how they intertwine with the necessity of AI governance platforms.
The Five Pillars of Trustworthy AI
The OECD's framework is built on five value-based principles:
1. Inclusive Growth and Sustainable Development: This principle emphasizes AI's role in enhancing human capabilities, fostering creativity, and promoting inclusion and equality. It's about harnessing AI for societal and environmental benefit.
2. Human-Centered Values and Fairness: AI should respect human rights, diversity, and democratic values. This includes privacy, data protection, and non-discrimination, ensuring that AI serves humanity and contributes to societal welfare.
3. Transparency and Explainability: This principle demands clarity about AI systems. The goal is to foster understanding and trust among users, enabling them to grasp the logic behind AI decisions and challenge outcomes if necessary.
4. Robustness, Security, and Safety: AI systems must be secure, safe, and resilient. They should function appropriately under various conditions, safeguarding users from undue risks.
5. Accountability: Lastly, AI actors must be responsible for their systems' functioning and adhere to these principles. This accountability is vital for trust in AI technologies.
Policy Recommendations for Trustworthy AI
Accompanying these principles, the OECD also provides policy recommendations for national governments and international cooperation. These include investing in AI R&D, shaping an enabling policy environment, building human capacity, and fostering international cooperation for trustworthy AI.
1. Investing in AI Research and Development:
- Governments are advised to consider long-term public investment and encourage private investment in AI R&D. This includes interdisciplinary efforts focusing on challenging technical issues and AI's social, legal, and ethical implications.
- Another aspect is investment in open datasets that are representative and respect privacy and data protection. This is crucial to create an AI research and development environment free of inappropriate bias and to enhance interoperability and standards.
2. Fostering a Digital Ecosystem for Trustworthy AI:
- The development of and access to a digital ecosystem that includes digital technologies, infrastructure, and mechanisms for sharing AI knowledge is emphasized.
- Governments should promote mechanisms like data trusts to support the safe, fair, legal, and ethical sharing of data.
3. Shaping an Enabling Policy Environment for AI:
- Policies should support a smooth transition from the R&D stage to the deployment and operation of trustworthy AI systems. Experimentation can be used to test and scale up AI systems in a controlled environment.
- Governments are advised to review and adapt policy and regulatory frameworks and assessment mechanisms as they apply to AI systems to encourage innovation and competition in the field of trustworthy AI.
4. Building Human Capacity and Preparing for Labor Market Transformation:
- This area focuses on the development of skills and capabilities to manage the transformation brought about by AI in the labor market. The emphasis is on preparing the workforce for the upcoming changes and opportunities presented by AI technologies.
5. International Cooperation for Trustworthy AI:
- Governments are encouraged to collaborate within the OECD and other global and regional forums to share AI knowledge.
- The development of multi-stakeholder, consensus-driven global technical standards for interoperable and trustworthy AI is recommended.
- Governments should also develop and use internationally comparable metrics to measure AI research, development, and deployment, and to assess progress in implementing these principles.
Connecting OECD AI Principles with AI Governance Platforms
For risk management professionals, the connection between these principles and AI governance platforms is clear. These platforms offer a centralized system to record and manage AI activities, ensuring compliance with these international standards. They provide tools for monitoring AI systems, ensuring transparency, robustness, and accountability. Moreover, they can aid in identifying and mitigating biases, ensuring inclusivity and fairness in AI applications.
Adopting OECD AI Standards: A Strategic Move
Incorporating OECD's standards isn't just about compliance; it's a strategic move towards responsible AI use. By aligning with these standards, companies not only demonstrate their commitment to ethical AI but also position themselves favorably in a market increasingly sensitive to how AI is used and governed.
The Future of AI Governance
As AI continues to evolve, so will the standards governing it. The OECD principles offer a robust foundation for responsible AI stewardship today and in the future. By leveraging AI governance platforms, companies can ensure they not only comply with these standards but also stay ahead in the rapidly advancing AI landscape.