Chapter 09 · Sustainability & AISustainability & AI

NIST AI Risk Management Framework

Meaning statusEstablishedSource recordDirect document linkedWhy these are different

Definition

A voluntary US framework (NIST AI 100-1, January 2023) for identifying and managing risks across the AI life cycle, organised around four functions: Govern, Map, Measure and Manage.

References

NISTAI Risk Management Framework 1.0 (NIST AI 100-1, 2023)

This source provides part of the technical or institutional basis for the definition.

NISTAI RMF Generative AI Profile (NIST AI 600-1, 2024)

This source supports the explanation of how the term is applied, measured or governed in practice.

Overview

What it means

Developed through multi-stakeholder consultation under a congressional mandate, the framework defines seven characteristics of trustworthy AI — including validity, safety, transparency, accountability, privacy and fairness — and links them to concrete actions. A Generative AI Profile (NIST AI 600-1, July 2024) added GenAI-specific risks, including confabulation and environmental impact.

How it is used

Organisations adopt the framework as the structure for AI risk programmes, inventories and assurance, and NIST continues to publish companion material, including the 2024 Generative AI Profile. The framework is referenced in US policy and procurement documents.

Why it matters

The AI RMF is the default risk vocabulary for AI in the US ecosystem and beyond. Its inclusion of AI's environmental footprint among GenAI risks moves that cost from externality toward managed risk.