Home / Insights / ISO/IEC 42001
Paper: Global, all sectors
ISO/IEC 42001 in plain language
ISO/IEC 42001 is the international standard for managing AI as a system, the way ISO/IEC 27001 does for information security. It is not a law. It gives your board, auditors and regulators one recognized way to see that your AI is under control.
Data Sentinels. Checked against the sources below on 13 November 2026. Not legal advice.
At a glance
Published
December 2023
What it is
Requirements for an AI management system: policies, roles, risk and impact assessment, controls and review
Who it is for
Any organization that builds, buys or uses AI
Certification
By an accredited body: a two-stage audit, yearly surveillance audits, recertification every three years
Related
ISO/IEC 42005 on AI impact assessment
What is inside
Context and scope
Which AI you build, buy or use, who cares about it, and where the management system starts and stops (clause 4).
Leadership and an AI policy
Top management owns it. There is a written AI policy and named roles (clause 5).
Risk and impact assessment
A repeatable way to assess AI risks and the impact of each AI system on people and society, with objectives to act on (clause 6).
People and records
Resources, competence, awareness and the documents that prove it (clause 7). Training is evidence, not a nice-to-have.
Control across the life cycle
AI is designed, tested, deployed, monitored and retired under control, with risk treatment and impact assessment applied (clause 8).
Check and improve
Monitoring, internal audit, management review, then corrective action and continual improvement (clauses 9 and 10).
The reference controls in Annex A
Annex A lists control objectives and controls you choose from and justify. They fall into nine areas:
A.2 Policies for AI
A.3 Internal organization
A.4 Resources for AI systems
A.5 Assessing the impacts of AI systems
A.6 The AI system life cycle
A.7 Data for AI systems
A.8 Information for interested parties
A.9 Use of AI systems
A.10 Third-party and customer relationships
What to do first
Three things you can start on Monday.
1. Set the scope: which AI systems and which parts of the organization are in.
2. Write the AI policy and start the AI register, with an owner for every system.
3. Run one impact assessment on your highest-risk system, so the method is tested before you scale it.
Sources
2. ISO, ISO/IEC 42001:2023 Artificial intelligence: Management system
This paper explains the standard. It does not reproduce it. Buy the text from ISO or your national standards body.
Not sure where you stand?
Take the 2-minute self-check. Eight questions, with your grade and red flags on screen.
Want it done for you?
Our Data & AI Governance and Risk work builds an AI management system aligned to ISO/IEC 42001: scope, policy, register, impact assessments and the evidence an auditor asks for.
Use the form below, or book a 30-minute call.
Not ready to talk yet?
Get Governance First: one email a month on what changed in the rules we track, and one thing to do about it.
One call is enough to know if it fits.
Send a request and we reply within one working day with times for a call.
In the United States? Use our US page.
Prefer email?
Rather pick a time now?
Book a 30-minute call. The call covers Data & AI Governance and Risk and the AI Academy.
What happens to your details
We use them to reply and prepare your call. Calls are recorded for quality and training. Read the privacy notice
