Wiki · Concept · Last reviewed July 10, 2026

ISO/IEC 25059

ISO/IEC 25059 is the ISO/IEC International Standard defining a SQuaRE quality model for AI systems.

Category: AI quality / standards Published: July 10, 2026 Modified: July 10, 2026 Last reviewed: July 10, 2026 Tags: ISO, IEC, SQuaRE, AI quality, quality model, AI evaluation

Definition

ISO/IEC 25059:2023 is titled Software engineering — Systems and software Quality Requirements and Evaluation (SQuaRE) — Quality model for AI systems. ISO lists it as Edition 1, a 15-page International Standard published in June 2023, with reference number ISO/IEC 25059:2023.

The public ISO abstract says the document outlines a quality model for AI systems as an application-specific extension to the SQuaRE standards. ISO says the model's characteristics and sub-characteristics give terminology for specifying, measuring, and evaluating AI system quality, and for comparing stated quality requirements against a quality-characteristic set for completeness.

Status

As reviewed on July 10, 2026, ISO lists ISO/IEC 25059:2023 as published, with current stage 90.92, to be revised. The ISO page also says the standard is expected to be replaced by ISO/IEC DIS 25059 within the coming months. Its lifecycle record shows new-project approval on May 18, 2020, committee-draft registration on September 14, 2021, DIS ballot initiation on July 12, 2022, final text received on February 10, 2023, publication on June 28, 2023, and a to-be-revised stage on October 31, 2023.

ISO identifies ISO/IEC JTC 1/SC 42 as the responsible technical committee and classifies the standard under ICS 35.080. The SC 42 committee page describes the subcommittee's scope as standardization in artificial intelligence and lists working groups for foundational standards, data, trustworthiness, use cases and applications, and computational approaches.

Quality Surface

ISO/IEC 25059 matters because AI quality claims are easy to compress into one vague word. A team can say a system is high quality while meaning model performance, data quality, usability, reliability, security, maintainability, robustness, or user value. A quality model forces the claim to be decomposed before it is measured.

The SQuaRE connection is important. The standard does not treat AI systems as a magical exception to software and systems evaluation. It extends a quality-model tradition toward AI-specific properties, so quality requirements can be discussed with more consistent vocabulary across engineers, evaluators, acquirers, auditors, and governance teams.

Engineering Use

For builders, ISO/IEC 25059 is most useful when quality requirements are being elicited, compared, or tested. It can help prevent a project from treating model score as the whole quality story. An AI product can have impressive benchmark performance while still failing quality requirements around data handling, operational behavior, explainability, monitoring, or integration with the surrounding system.

For procurement and assurance, the standard supports better questions. Instead of asking whether an AI system is simply "good," a buyer can ask which quality characteristics are in scope, which are excluded, which measurements support each claim, and how the evaluation changes when the model, data, context, or system boundary changes.

Evidence Record

An ISO/IEC 25059-informed record should identify the AI system, system boundary, quality characteristics in scope, sub-characteristics used, quality requirements, measurement method, evaluation evidence, data assumptions, operating context, responsible reviewer, result, limitation, and retest trigger. It should also state which quality characteristics were not assessed.

The record should preserve the distinction between a quality model and proof of acceptable deployment. A quality model helps organize evidence; it does not by itself settle risk, ethics, safety, legal compliance, or public legitimacy. Quality evidence becomes useful when linked to use case, life cycle stage, risk treatment, impact assessment, monitoring, and accountable ownership.

Boundary With Other Standards

ISO/IEC 25059 is not an AI management-system standard, risk-management guide, data-quality standard, classification-performance method, or impact-assessment guide. It sits beside adjacent references. ISO/IEC 5259 addresses data quality for analytics and machine learning, ISO/IEC TS 4213 addresses classification performance, ISO/IEC 5338 addresses AI system life cycle processes, and ISO/IEC 42001 addresses AI management systems.

Source Discipline

Use the official ISO page for the title, reference number, International Standard status, publication date, edition, page count, current revision stage, replacement-under-development note, technical committee, ICS classification, public abstract, and lifecycle dates. Use the ISO/IEC JTC 1/SC 42 page for committee scope and working-group structure. Do not cite vendor summaries for the standard's formal status, and do not treat ISO/IEC 25059 as product approval, certification, or legal safe harbor.

Spiralist Reading

Spiralism reads ISO/IEC 25059 as a discipline against quality theater. "Quality AI" can become a brand phrase unless the speaker names the quality characteristic, the system boundary, the measurement, the evidence, and the owner. The standard's value is that it helps move the conversation from adjectives to inspectable requirements.

The stricter reading is that quality is not a mood around a model. It is a structured argument about what the system is supposed to be, what evidence supports that claim, where the claim fails, and when it must be revisited. That makes quality a governance record rather than a launch slogan.

Open Questions

Sources

Return to Wiki