Standards Development

Scope

This document provides measurement and guidance for quality evaluation of artificial intelligence (AI) systems using AI system quality models defined in ISO/IEC 25059. It contains a set of quality measures and guidance for AI system quality models.

The document is applicable to all types of organizations engaged in the development and use of AI systems.

Purpose

The purpose of this standard is to provide guidance for the quality evaluation of artificial intelligence (AI) systems in accordance with the revised AI system quality models. In addition, this standard provides quality measures for AI system quality models.

(Guidance)

ISO/IEC 25059, the foundational standard for this document, includes revised quality characteristics for the AI product quality model and the quality-in-use model, and the AI service quality model is added. Therefore, guidance is needed to support quality evaluation based on the revised and added quality models.

(Measurement)

Due to the quality characteristics of AI systems, the quality measures provided by the existing SQuaRE series (e.g. ISO/IEC 25022, ISO/IEC 25023, ISO/IEC TS 25025) for traditional systems and software may not always be directly applicable to the quality evaluation of AI systems. Furthermore, the SQuaRE series does not provide quality measures for the modified or newly introduced quality characteristics of AI systems, making it challenging to perform quantitative quality evaluation for these specific quality characteristics.

To address this, measurement for the AI system quality model and corresponding revised guidelines are required.

In addition, since the proposed standard provides both guidance and measurement within a single document, the technical content is sufficiently stable and well-understood to be developed as an International Standard(IS), rather than as a Technical Specification(TS).