AI governance frameworks: NIST, ISO 42001 and the EU AI Act for executives
You do not need to adopt every framework. You need to know which ones are voluntary, which one is law where you operate, and how to build one set of controls that satisfies them.
An AI governance framework is a structured way to make sure AI is used lawfully, safely and to good effect. Four references matter most to executives: the NIST AI Risk Management Framework, a voluntary United States framework; ISO/IEC 42001, an international management-system standard a company can be certified against; the EU AI Act, a law with obligations by risk level; and the OECD AI Principles, the values most of the others build on.
NIST
ISO
European Commission
OECD
Microsoft 365
Anthropic
How to read this chart
The scores are our planning judgement for a typical small or mid-size company: they show where effort or risk tends to concentrate, not measured data. Replace them with your own once you have looked at your company.
What does the NIST AI Risk Management Framework ask for?
It organises AI risk work into four functions. Govern sets the culture, roles and policies. Map establishes the context and the risks of a given use. Measure assesses and tracks those risks. Manage acts on them.
- It is voluntary and not tied to any one sector or size of company.
- A companion Generative AI Profile applies it to generative AI.
- Most small companies can apply it as a checklist per use case rather than as a program.
When is ISO/IEC 42001 worth pursuing?
When customers, partners or regulators want independent assurance that AI is managed. It is a management-system standard in the same family as the information-security standard many companies already hold, and it can be audited and certified.
- Consider it if you sell AI-enabled products or services to larger companies.
- It builds on an inventory, a policy and risk reviews; do those first.
- Certification is a decision about market need, not a first step.
What do the EU AI Act's dates mean for a company outside the EU?
The Act can apply to companies outside the EU that place AI systems on the EU market or whose systems' output is used there. It applies in stages, and the high-risk dates were moved in 2026.
- Since 2 February 2025: banned practices and the AI literacy duty apply.
- Since 2 August 2025: the rules for general-purpose AI models and governance apply.
- From 2 December 2027: the rules for high-risk uses in listed sensitive areas, such as employment and education, apply.
- From 2 August 2028: the rules for high-risk AI built into regulated products apply.
These dates are from the European Commission's AI Act page as reviewed on the date shown. The timeline changed in 2026; check the Commission's page and your counsel before relying on any date.
How do you build one set of controls for all of them?
Start from what they have in common. Every one of them expects you to know where AI is used, who is accountable, what the risks are and what you do when something goes wrong.
- Keep an inventory of AI uses with an owner for each.
- Write a use policy and a list of approved tools.
- Review each use case for risk before it goes live, in proportion to its impact.
- Check vendor terms on data use, retention and training.
- Monitor live uses and keep an incident process.
- Map those controls to the frameworks that apply to you.
Questions executives ask about this
Is the NIST AI RMF mandatory?
No. It is a voluntary framework. Some contracts, sector regulators and state rules refer to it, so check what applies to your company.
Does the EU AI Act apply to a United States company?
It can, if the company places AI systems on the EU market or the output of its systems is used in the EU. Whether it applies to you is a legal question for your counsel.
Which framework should a mid-size company start with?
Use the NIST AI RMF as the working structure, because it is free and practical. Add ISO/IEC 42001 if customers ask for certification, and the EU AI Act if you have EU exposure.
Is this legal advice?
No. It is a plain-language summary for executives. Have your counsel confirm what applies to your company.
Primary sources to check against
AI Risk Management Framework, AI RMF 1.0 (National Institute of Standards and Technology)
AI RMF Generative AI Profile, NIST AI 600-1 (National Institute of Standards and Technology)
ISO/IEC 42001:2023, AI management systems (International Organization for Standardization)
AI Act: the regulatory framework for AI and its application timeline (European Commission)
OECD AI Principles (OECD.AI Policy Observatory)
Frameworks, laws and product details change. Confirm the current text with each publisher and with your own counsel. Consultants in a Box is an independent consultancy; this page is not published or endorsed by any organisation named. Names and logos are trademarks of their owners.
Related guides
AI acceptable use policy
Staff are already using AI. A policy that fits on two pages, names the approved tools and says plainly what may not be pasted into them does more than a long one nobody reads.
Read the guideBoard oversight of AI
A board does not need to understand how a model works. It needs to know where the company uses AI, what could go wrong, who is accountable and whether the investment is paying back.
Read the guideAI operating model
An operating model answers three questions: who chooses what gets built, who builds and runs it, and who answers for the result. Without it, every department buys its own tool and nobody owns the risk.
Read the guideAll AI for executives guides
The full leadership playbook, from readiness to agents.
See the hubWant to talk this through for your company?
Book a call and we will go through where you are, what to decide first and who should own it. Or start with the free business analysis.
