Artificial Intelligence — Concepts and Terminology

ISO/IEC 22989:2022(E), titled "Artificial Intelligence — Concepts and Terminology," is a foundational standard published by ISO and IEC. It establishes a globally agreed-upon vocabulary for the AI field, defining key concepts such as "AI system," "machine learning," and "trustworthiness." In enterprise risk management, it serves as a common language, enabling precise risk communication. By providing a clear definition of "algorithmic bias," it helps organizations align with frameworks like the NIST AI Risk Management Framework (AI RMF 1.0) and prepare for regulations like the EU AI Act. It is a prerequisite for building robust AI governance systems like ISO/IEC 42001, ensuring all stakeholders share a common understanding of AI-related risks.

Enterprises can apply ISO/IEC 22989 in risk management through a three-step process. First, **Internal Alignment and Training**: Map internal AI terms to the standard's definitions and train all relevant teams. Second, **Integration into Risk Frameworks**: Update risk assessment processes based on ISO 31000 to include specific AI risks defined by the standard (e.g., model drift). Third, **Vendor and Contract Management**: Use the standard's terminology in RFPs and contracts with AI vendors to clarify requirements. For example, a global bank can use the standard to unify AI model risk terminology, which can measurably improve regulatory audit pass rates by over 15% and reduce ambiguity in third-party AI solution procurement.

Taiwan enterprises face three key challenges. First, **Lack of Local Regulation**: Without a dedicated AI law like the EU AI Act, the immediate compliance incentive is low. Second, **Limited Expertise**: SMEs often lack personnel for AI governance. Third, **Cultural Gaps**: A disconnect exists between technical and management teams. To overcome this, enterprises should proactively align with global standards like the NIST AI RMF. A phased implementation, starting with a critical project and supported by external consultants, can address resource constraints. Establishing a cross-functional AI governance committee that uses the standard as its official language can bridge the cultural gap, with the first priority being the creation of a unified internal glossary.

Winners Consulting specializes in ISO/IEC 22989:2022(E) for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact