Types of standards ISO has for AI

ISO has developed a growing ecosystem of AI-related standards to address different aspects of artificial intelligence, from terminology and technical design to risk management and governance. These standards are designed to work together, with ISO 42001 acting as the central management framework. ISO/IEC 22989 – AI concepts and terminology This standard defines key AI concepts and terminology to ensure consistency and shared understanding across organizations, regulators, and technical teams. It helps prevent misinterpretation by ensuring everyone uses the same definitions when discussing AI systems. ISO/IEC 23053 – Machine learning framework ISO/IEC 23053 provides guidance on designing and implementing machine learning systems, including data preparation, training, validation, and deployment. It supports technical teams by offering a structured approach to building ML-based AI systems. ISO/IEC 23894 – AI risk management This standard focuses specifically on identifying and managing risks associated with AI systems. It complements ISO 42001 by providing more in-depth guidance on AI-specific risk scenarios. ISO/IEC 38507 – Governance of IT with AI considerations ISO/IEC 38507 helps leadership teams govern AI within broader IT and organizational governance frameworks. It focuses on oversight, accountability, and strategic alignment. ISO/IEC 42001 – AI management systems ISO 42001 ties everything together by providing a management system that governs how AI is planned, deployed, monitored, and improved across the organization. It acts as the governing layer, while other standards provide technical and conceptual support.