NIST AI Risk Management Framework: A Plain-English Guide for Business Owners and Boards
A 48-page US government framework that has become a common language for managing the risks of AI. What it says, what it leaves out, and how a business that buys AI can use it without a data science team.
- Source:
- National Institute of Standards and Technology (NIST), US Department of Commerce
- Original report:
- 26 January 2023
- Edition:
- Version 1.0, January 2023 (NIST AI 100-1)
- On LegalLab:
- 19 October 2026
The report in one page
The short version
The Artificial Intelligence Risk Management Framework (AI RMF 1.0), published by the US National Institute of Standards and Technology (NIST) in January 2023, is a short, voluntary guide to managing the risks of AI systems. It is not a law. It gives an organisation a common vocabulary and a set of outcomes to aim for, whether it builds AI or only buys it (pp. 1-2).
Its core idea is that AI risk is different from ordinary software risk. AI systems learn from data that can change, behave in ways that are hard to explain, and are shaped by the people and settings around them, which NIST calls "socio-technical" (p. 1). So the framework asks an organisation to do four things, which it names GOVERN, MAP, MEASURE and MANAGE (p. 20):
- Govern: set the culture, policies, roles and accountability for AI risk. This runs through everything else.
- Map: understand the context of an AI system, what it is for, who it affects and what could go wrong.
- Measure: test and track those risks with suitable methods and independent review.
- Manage: decide what to do about them, plan for incidents and monitor after launch.
The framework also lists seven characteristics of a trustworthy AI system: valid and reliable, safe, secure and resilient, accountable and transparent, explainable and interpretable, privacy-enhanced, and fair with harmful bias managed (p. 12). NIST is clear that these pull against each other and that no formula settles the trade-offs. Context and human judgement do (pp. 12-13).
Three points matter for a business reader. The framework does not set a risk tolerance for you (p. 7). It expects AI risk to sit inside the organisation's wider risk management, alongside cybersecurity and privacy (p. 8). And it says plainly that a framework alone will not change behaviour: that needs commitment from senior leaders (p. 9).
The practical message is to treat AI like any other material business risk: know where it is used, say who is responsible, test it before and after launch, and be ready to switch it off.
Printing is available on paid plans – legallab.co.in