AI standards

AI Standards

There are many ISO standards regarding AI and there is an active development on the following standards related to AI.

Some standards are related to AI standards in a non-medical environment and additional standards are available for medical device development. Depending on the device and implementation some of these standards will apply to your solution.

Besides the most common standards listed below there are additional standards available that might suit or apply to yourr AI solution.

Management system for AI development

ISO 42001, which establishes a management system for responsible AI development and implementation, providing a framework for governance, ethics, transparency, and risk management throughout the AI lifecycle.

ISO 42003: Guidance on implementation of 42001

ISO 22989: AI concepts and terminology

ISO/IEC 23053: Machine learning systems

ISO 42005: methodologies for AI impact assessments

ISO 23894: AI risk management

Generic standards for software testing

ISO/IEC E 29119 defines an internationally agreed set of standards for software testing that can be used by any organization when performing any form of software testing.

  1. ISO 29119-1 introduces software testing concepts.
  2. ISO 29119-2 comprises test process descriptions that define the software test processes at the
    organizational level, test management level and dynamic test levels.
  3. ISO 29119-3 defines software test documentation.
  4. ISO 29119-4 defines test design techniques.
  5. ISO 29119-5 provides keyworrd-driven testing.
  6. ISO 29119-6 defines the guidelines for the use in agile systems

Processes and templates for work reviews

  1. ISO 20246 defines processes and templates for work product reviews, including inspections,
    walkthroughs and technical reviews.

Guidance for testing of AI components

The ISO 42119 series is to provide requirements and guidance on the testing of AI components and systems

  1. TS 42119-3 describes approaches and provides guidance on processes for the verification and
    validation analysis of AI systems
  2. TS 42119-7 provides technology-agnostic guidance for conducting red teaming assessments on
    AI systems
  3. TS 42119-8 provides definitions, concepts, requirements and guidance related to assessing
    prompt-based text-to-text AI systems that utilize generative AI.