OECD's values-based principles

The OECD's values-based principles are a set of five core tenets from the 'Recommendation of the Council on Artificial Intelligence,' adopted in May 2019. They provide a global standard for the responsible stewardship of trustworthy AI. The principles are: (1) inclusive growth, sustainable development, and well-being; (2) human-centred values and fairness; (3) transparency and explainability; (4) robustness, security, and safety; and (5) accountability. While not legally binding, they function as influential 'soft law,' shaping national AI strategies and international standards like ISO/IEC 42001:2023 (AI Management System). In enterprise risk management, they serve as a high-level ethical framework to guide the development and deployment of AI systems, ensuring alignment with human rights and societal values.

Enterprises can operationalize the OECD principles through a structured approach: 1. **Establish Governance:** Form a cross-functional AI ethics committee to create internal policies based on the principles, fulfilling the leadership and accountability requirements outlined in standards like ISO/IEC 42001. 2. **Conduct AI Impact Assessments (AIA):** For high-risk applications, mandate AIAs, similar to GDPR's DPIAs. This involves systematically identifying and mitigating risks related to fairness, transparency, and safety, using frameworks like the NIST AI Risk Management Framework (AI 100-1) as a guide. 3. **Implement Technical & Monitoring Tools:** Deploy eXplainable AI (XAI) tools (e.g., SHAP, LIME) to enhance model transparency. Establish continuous monitoring dashboards to track fairness metrics and model drift. A global bank that implemented this process reduced gender bias in its loan-approval AI by 35%, improving both regulatory compliance and customer trust.

Taiwan enterprises face three primary challenges: 1. **Regulatory Ambiguity:** Without a dedicated AI law, translating high-level principles like 'transparency' into concrete actions that comply with existing regulations, such as the Personal Data Protection Act, is difficult. 2. **Resource Constraints for SMEs:** The cost of establishing dedicated ethics boards and acquiring specialized tools for bias detection and explainability poses a significant barrier for small and medium-sized enterprises (SMEs), which form the backbone of Taiwan's economy. 3. **Data Bias and Quality:** Local datasets may contain inherent historical or societal biases, which can lead AI models to violate the 'fairness' principle. Addressing this requires advanced data governance and technical expertise. **Solutions:** Enterprises should adopt a risk-based approach, prioritizing high-impact systems, leverage open-source tools to manage costs, and partner with expert consultancies to build a foundational governance framework efficiently.

Winners Consulting specializes in OECD's values-based principles for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact