Definition of ISO 42001
ISO 42001 is the common name of the international standard ISO/IEC 42001:2023, entitled “Information technology, Artificial intelligence, Management system”. According to the summary published by the AFNOR store, it specifies the requirements and guidance for establishing, implementing, maintaining and continually improving an AI management system. The same listing dates it to December 2023 and shows it as in force. The store also lists the French standard NF EN ISO/IEC 42001 (March 2026), which adopts the European standard EN ISO/IEC 42001:2026.
The text is paid for: this entry does not describe its clauses. It relies on public pages, consulted on 12 November 2026: the AFNOR store, the AFNOR Certification page and the preview of the preliminary pages published by VDE Verlag.
What an AI management system is for
According to its scope, the standard aims to help any organisation that provides or uses products or services that use AI systems to develop, provide or use these systems responsibly.
The introduction cites three considerations specific to AI: automated decisions that are sometimes not very transparent, systems designed through machine learning rather than hand-written logic, and behaviours that change during use. It adds that an organisation conforming to the requirements can produce evidence of the way it fulfils its role.
What the table of contents of the standard shows
The headings below were read in English in the table of contents of the preview.
| Part | Heading |
|---|---|
| Clauses 4 to 10 | Context of the organisation, leadership, planning, support, operation, performance evaluation, improvement |
| AI-specific subclauses | AI policy, AI risk assessment, treatment of these risks, AI system impact assessment |
| Annexes A and B (normative) | Reference objectives and controls; implementation guidance for AI controls |
| Annexes C and D (informative) | Potential AI-related objectives and risk sources; use of the management system across domains or sectors |
According to the introduction, the standard applies the harmonised structure of management system standards, which makes consistency with those for quality, security or privacy easier.
ISO 42001, AI Act and certification: not to be confused
ISO 42001 is a standard, whose foreword refers to the voluntary nature of standards. The AI Act, Regulation (EU) 2024/1689, is a legal text. Its consolidated version of 27 July 2026 does not cite the standard. Its Article 40 attaches a presumption of conformity to harmonised standards whose references are published in the Official Journal of the European Union. None of the sources consulted places ISO/IEC 42001, or its European adoption, among these standards.
Certification exists: AFNOR Certification presents it and describes the standard as certifiable. The AFNOR store also lists ISO/IEC 42006:2025 (July 2025), whose title covers bodies that audit and certify these management systems. The object certified is therefore an organisation’s management system. The conformity of a high-risk AI system is assessed according to the procedures of the regulation (Article 43).
The NIST AI risk management framework, another voluntary framework, is presented in our article on implementing AI governance.
Example: a mutual insurer structuring its AI practices
Take a mutual insurer that uses several purchased AI systems and is developing one. It already runs an information security management system, with internal audits and management reviews. Its management takes ISO 42001 as a reference: it acquires the text, sets a scope, compares its practices with the requirements and reuses these audits and reviews. Whether or not to seek certification remains a separate decision.
The place of such a framework alongside the register and classification of systems is covered on our page on AI governance.