NIST AI Risk Management Framework 1.0

Publisher: National Institute of Standards and Technology Version: 1.0

A voluntary framework for managing risks throughout the AI lifecycle. Organized around four core functions: Govern, Map, Measure, and Manage. Includes the Generative AI Profile (AI 600-1) for foundation model risks.

Domains: Tier:

ID	Name	Description	Domains
GV	GOVERN — Policies, Accountability & Culture	The GOVERN function establishes and maintains the organizational structures, policies, processes, and culture necessa...	govern
GV-1	Policies, processes, and practices	AI policies, processes, procedures, and practices are in place across the organization to map, measure, manage, and g...	govern
GV-2	Accountability structures	Roles, responsibilities, and lines of communication related to AI risk management are established with clear accounta...	govern
GV-3	Diversity, equity, inclusion, and accessibility	Organizational teams building, deploying, and using AI systems reflect diversity, and are proactive in addressing har...	govern
GV-4	Organizational culture and commitment	Organizational culture and leadership foster responsible stewardship of trustworthy AI aligned with societal values. ...	govern
GV-5	Stakeholder engagement	Organizational practices are in place to enable AI deployment and ongoing use with input from affected communities an...	govern
GV-6	Supply chain risk management	AI risks and benefits from third-party resources are regularly monitored, and risk controls are applied and documente...	govern
MP	MAP — Context & Risk Identification	The MAP function establishes the context for AI risk management by identifying the intended purpose, deployment conte...	build
MP-1	Context establishment	Legal, regulatory, and societal contexts of AI system deployment are identified and documented, including norms and e...	build
MP-2	Categorization and risk tiering	AI systems are categorized based on their intended use, beneficiaries, and potential for harm, enabling risk-proporti...	build
MP-3	AI capabilities and limitations	AI system capabilities, intended purposes, context of use, and known limitations are documented. Every AI system has ...	build
MP-4	Risk identification and analysis	AI risks and benefits are identified, assessed, prioritized, and documented covering technical, societal, and ethical...	build
MP-5	Impact assessment	AI system impacts on individuals, groups, communities, organizations, and society are identified and assessed. Impact...	build
MS	MEASURE — Assessment, Metrics & Testing	The MEASURE function employs quantitative and qualitative methods to assess AI system trustworthiness characteristics...	monitor
MS-1	Measurement approach	Appropriate methods and metrics are identified and applied to measure AI system performance, fairness, safety, and ot...	monitor
MS-2	Testing and validation	AI systems undergo testing and validation including adversarial testing, bias testing, and red-teaming appropriate to...	monitor
MS-3	Competency and expertise	AI system measurement activities are performed by individuals and teams with appropriate domain knowledge, technical ...	monitor
MS-4	External inputs and validation	Measurement processes incorporate external perspectives, independent testing, and stakeholder feedback to validate AI...	monitor
MG	MANAGE — Risk Response & Communication	The MANAGE function allocates resources, implements controls, deploys mitigations, and establishes ongoing monitoring...	deploy monitor
MG-1	Purpose evaluation	AI systems are evaluated to determine whether their intended purpose, use cases, and deployment context are appropria...	deploy monitor
MG-2	Risk and benefit balancing	AI system risks and benefits are balanced and managed based on expected impact, with risk tolerance aligned to organi...	deploy monitor
MG-3	Lifecycle monitoring	AI systems are monitored throughout their lifecycle to detect performance degradation, drift, emerging risks, and uni...	deploy monitor
MG-4	TEVV (Test, Evaluation, Verification, and Validation)	TEVV processes are implemented and iterated throughout the AI lifecycle to ensure systems function as intended and al...	deploy monitor
AI 600-1	Generative AI Profile	NIST AI 600-1 is a companion document to the AI RMF that addresses risks unique to generative AI systems including fo...	secure build
GAI-1	CBRN Information and Capabilities	Risk that generative AI could lower barriers to development or use of chemical, biological, radiological, or nuclear ...	secure build
GAI-2	Confabulation (Hallucination)	Risk of generative AI producing false, fabricated, or inconsistent outputs presented as factual information with high...	secure build
GAI-3	Data Privacy	Risk of generative AI exposing, inferring, or memorizing personal or sensitive information from training data. Large ...	secure build
GAI-4	Environmental Impact	Risk of significant energy consumption and carbon emissions from training and operating large generative AI models. T...	secure build
GAI-5	Homogenization of Content and Perspectives	Risk that widespread generative AI use reduces diversity of content, ideas, and cultural expressions. When everyone u...	secure build
GAI-6	Human-AI Configuration	Risk of inappropriate reliance on generative AI, automation bias, or degradation of human skills and judgment. When h...	secure build
GAI-7	Information Integrity	Risk of generative AI enabling misinformation, disinformation, deepfakes, and manipulation of information ecosystems ...	secure build
GAI-8	Information Security	Risk of adversarial attacks targeting generative AI systems including prompt injection, jailbreaking, model inversion...	secure build
GAI-9	Intellectual Property	Risk of generative AI infringing copyright, reproducing protected works, or creating unclear IP ownership of AI-gener...	secure build
GAI-10	Obscene, Degrading, and Abusive Content	Risk of generative AI producing violent, sexual, hateful, or otherwise harmful content. Without safety guardrails, ge...	secure build
GAI-11	Toxicity, Bias, and Homogenization	Risk of generative AI amplifying stereotypes, producing biased outputs, or exhibiting toxic language patterns. Langua...	secure build
GAI-12	Value Chain and Component Integration	Risk from dependencies on third-party foundation models, APIs, plugins, or components with unknown properties or vuln...	secure build