ISO/IEC 22989: Artificial intelligence — concepts and terminology
ISO/IEC 22989:2022 establishes a common vocabulary for artificial intelligence (AI). It defines core concepts and terms for AI systems, data, lifecycle stages, roles, and AI properties, to enable consistent communication across stakeholders and to underpin other AI standards from ISO/IEC JTC 1/SC 42.
Scope and intent
- Defines foundational AI concepts (e.g., AI system, model, algorithm, dataset; training/validation/testing data; lifecycle; deployment and operation)
- Clarifies human roles and oversight (human-in-the-loop/on-the-loop/over-the-loop) and organizational responsibilities
- Introduces terminology for key AI properties such as transparency, explainability, robustness, reliability, resilience, safety, security, privacy, and risk
- Serves as a reference for related AI standards (risk management, trustworthiness, management systems, data quality, testing)
Position in the SC 42 family
- Terminology baseline used by:
- ISO/IEC 23894 — AI risk management
- ISO/IEC 23053 — Framework for AI systems using machine learning
- ISO/IEC 42001 — Artificial Intelligence Management System (AIMS)
- ISO/IEC 24028 — Overview of trustworthiness in AI
- ISO/IEC 5259 (series) — Data quality for analytics and ML
- ISO/IEC/IEEE 29119-11 — Testing of AI-based systems
Quality Attributes Addressed (via terminology and concept coverage)
| Attribute | How it is addressed or framed in 22989 |
|---|---|
| Reliability | Terms for performance consistency and dependable behavior across lifecycle phases |
| Safety | Concepts relating to harm, hazard, and safe operation of AI systems |
| Security | Terminology linking security properties (confidentiality, integrity, availability) to AI contexts |
| Robustness | Definitions around robustness to perturbations, uncertainty, and dataset shift |
| Resilience | Concepts for recovery and continued operation under adverse conditions |
| Transparency | Shared language for making AI system capabilities and limitations visible |
| Explainability | Definitions for explainability/interpretability to support understanding of outputs |
| Accountability | Roles and responsibilities, human oversight, assurance concepts |
| Fairness / Bias mitigation | Terms for bias, fairness, and mitigation approaches |
| Privacy | Concepts for data protection in AI lifecycles |
| Data quality | Dataset, labeling, quality characteristics across training/validation/testing |
| Usability / human factors | Human-in/on/over-the-loop, human oversight terminology |
| Maintainability | Lifecycle and change-related terms that support maintainable operation |
| Traceability | Terminology for artifacts, provenance, and evidence across the lifecycle |
References
- ISO/IEC 22989:2022 — Artificial intelligence concepts and terminology: https://www.iso.org/standard/74296.html
- ISO Online Browsing Platform (standard preview): https://www.iso.org/obp/ui/#iso:std:iso-iec:22989:ed-1:v1:en
- ISO/IEC JTC 1/SC 42 (Artificial intelligence) committee: https://www.iso.org/committee/6794475.html
- ISO/IEC 42001:2023 — Artificial Intelligence Management System (AIMS): https://www.iso.org/standard/81230.html
- NIST AI RMF Crosswalks (alignment with ISO/IEC standards): https://www.nist.gov/itl/ai-risk-management-framework/ai-rmf-crosswalks