NIST AI Risk Management Framework (AI RMF)

The NIST AI Risk Management Framework is the U.S. National Institute of Standards and Technology’s voluntary framework for managing AI risk, organized around four functions — Govern, Map, Measure, Manage — to help organizations build trustworthy AI.

The framework

AI RMF 1.0 was released January 26, 2023, “intended for voluntary use” to improve trustworthiness considerations across the AI lifecycle — design, development, deployment, use, and evaluation nist-ai-rmf-overview. NIST ships it with a companion AI RMF Playbook (suggested actions per function), a roadmap, and crosswalks mapping it to other frameworks.

Four core functions. Govern (the cross-cutting culture, policies, and accountability that the other three run inside) · Map (establish the context and identify risks) · Measure (analyze, assess, and track the mapped risks) · Manage (prioritize and act on them).

Seven trustworthiness characteristics the framework organizes around: valid and reliable · safe · secure and resilient · accountable and transparent · explainable and interpretable · privacy-enhanced · fair, with harmful bias managed nist-ai-rmf-overview.

Generative-AI Profile. On July 26, 2024 NIST released NIST-AI-600-1, the Generative Artificial Intelligence Profile — a use-case profile extending the core framework to risks specific to generative systems nist-ai-rmf-overview.

Role in this wiki (so far)

A voluntary counterpart to the binding eu-ai-act: nemko-ai-governance-services aligns to the NIST AI RMF as one of its reference frameworks. Sits with iso-iec-42001 on the organizational side of ai-governance, operationalizing risk-based-regulation inside a firm. Its voluntary nature makes it a natural fit for soft-law regimes like Israel’s responsible-innovation.

Related

ai-governance · iso-iec-42001 · eu-ai-act · risk-based-regulation · oecd-ai-principles · nemko-ai-governance-services · nist-ai-rmf-overview