Reference number
ISO/IEC 25059:2023
International Standard
ISO/IEC 25059:2023
Software engineering — Systems and software Quality Requirements and Evaluation (SQuaRE) — Quality model for AI systems
Edition 1
2023-06
Preview
ISO/IEC 25059:2023
80655
Indisponible en français
Publiée (Edition 1, 2023)

ISO/IEC 25059:2023

ISO/IEC 25059:2023
80655
Langue
Format
CHF 98
Convertir les francs suisses (CHF) dans une autre devise

What is ISO/IEC 25059?

ISO/IEC 25059 provides a dedicated quality model for artificial intelligence (AI) systems as an extension of the SQuaRE (Software product Quality Requirements and Evaluation) framework. It defines quality characteristics and sub-characteristics tailored to AI-specific behaviours, offering a standardised way to specify, assess and ensure the quality of AI systems.

Why is ISO/IEC 25059 important?

Traditional software quality models are not enough for AI systems, which often behave unpredictably, adapt over time and make probabilistic decisions. ISO/IEC 25059 bridges this gap by introducing a quality model that reflects the unique properties of AI — including trustworthiness, data dependency and dynamic learning. By using this model, organisations can set and evaluate AI system requirements more clearly and consistently across different domains and tasks.

Benefits

  • Provides a dedicated quality model specific to AI systems
  • Enables consistent specification and evaluation of AI quality attributes
  • Enhances trust and reliability of AI-enabled products and services
  • Supports alignment with broader SQuaRE and data quality standards
  • Facilitates integration of AI quality considerations in system development

 

FAQ

Any organisation that develops, integrates or evaluates AI systems and wants to ensure they meet quality expectations.

ISO/IEC 25059 extends the ISO/IEC 25010 model by introducing characteristics specific to AI systems, while maintaining consistency with the overall SQuaRE framework.

The model is applicable across all AI systems, including those based on machine learning, probabilistic reasoning or other adaptive approaches.

Informations générales

  •  : Publiée
     : 2023-06
    : Norme internationale à réviser [90.92]
  •  : 1
  • ISO/IEC JTC 1/SC 42
    35.080 
  • RSS mises à jour

Vous avez une question?

Consulter notre Aide et assistance