Artificial intelligence — Quality evaluation guidelines for AI systems

Scope

This document provides guidelines for evaluation of AI systems using an AI system quality model. The document is applicable to all types of organizations engaged in the development and the use of artificial intelligence. PurposeAI systems can be very challenging to evaluate. Concurrently, the impact of poor AI systems’ quality is considerable as they are often developed to facilitate the automation of critical actions and decisions. The purpose of this technical specification is to guide AI developers through proper quality evaluation of their AI systems. This document does not state exact metrics and thresholds as they vary depending on the nature of each system. Instead, it specifies a comprehensive set of guidelines that cover relevant facets of AI system’s quality for successful quality evaluation. Testing is in scope as far as each characteristics and sub-characteristics are verified by testing strategies, but details of testing methods and metrics is covered elsewhere, for example in ISO/IEC/IEEE 29119 Series. © British Standards Institution 2022

Purpose

AI systems can be very challenging to evaluate. Concurrently, the impact of poor AI systems’ quality is considerable as they are often developed to facilitate the automation of critical actions and decisions.

The purpose of this technical specification is to guide AI developers through proper quality evaluation of their AI systems. This document does not state exact metrics and thresholds as they vary depending on the nature of each system. Instead, it specifies a comprehensive set of guidelines that cover relevant facets of AI system’s quality for successful quality evaluation.

Testing is in scope as far as each characteristics and sub-characteristics are verified by testing strategies, but details of testing methods and metrics is covered elsewhere, for example in ISO/IEC/IEEE 29119 Series. ©ISO/IEC 2022. All rights reserved.

External Links