NIST AI Risk Management Framework: AI RMF 1.0 Guide | growGRC
NIST AI Risk Management Framework: A Practical Guide to AI RMF 1.0
Practical guide to NIST AI RMF 1.0: Govern, Map, Measure and Manage, lifecycle, GRC integration, Playbook, crosswalks and Generative AI Profile.
Published: 8/23/2026Updated: 8/23/2026By growGRC Editorial Team
Executive Summary
NIST AI RMF 1.0 is a voluntary framework for managing AI risk through Govern, Map, Measure and Manage. This guide explains the Core, lifecycle, Playbook, GRC integration, crosswalks and current revision status.
Key Takeaways
✓NIST AI RMF 1.0 was published January 26, 2023.
✓The AI RMF Core has four functions: Govern, Map, Measure and Manage.
✓The functions are continuous and are not a linear checklist.
✓The Playbook provides voluntary suggested actions and references.
✓NIST currently states that AI RMF 1.0 is being revised.
NIST AI Risk Management Framework: A Practical Guide to AI RMF 1.0
The NIST AI Risk Management Framework (AI RMF) is a voluntary framework developed by the National Institute of Standards and Technology to help organizations that design, develop, deploy, or use artificial intelligence manage AI risks and promote trustworthy and responsible AI. NIST published AI RMF 1.0 on January 26, 2023. citeturn0search0turn0search3
NIST describes AI RMF 1.0 as voluntary, rights-preserving, non-sector specific, and use-case agnostic. It is intended to be flexible enough for organizations of different sizes and sectors and to be operationalized at different levels of capacity. citeturn0search0
Current status: NIST states that AI RMF 1.0 is being revised. The current published framework remains AI RMF 1.0, while NIST's companion Playbook is expected to be updated after the framework is revised. citeturn0search6turn0search2
What is the NIST AI Risk Management Framework?
NIST AI RMF 1.0 provides a common structure for thinking about AI risk throughout the AI lifecycle.
It is not a technology product, certification, or prescriptive implementation methodology. Instead, it provides outcomes, categories, subcategories, and actions that organizations can adapt to their own context.
The framework is built around two complementary ideas:
1. Framing AI risk — understanding the context, potential impacts, benefits, harms, and challenges that influence AI risk.
2. Managing AI risk through the AI RMF Core — using Govern, Map, Measure, and Manage to organize risk management activities.
The Core is designed to support dialogue, understanding, and practical activities for managing AI risks and developing trustworthy AI systems. citeturn0search5turn0search29
AI systems can introduce risks that are difficult to manage using traditional software or risk processes alone. NIST developed AI RMF to provide organizations with a practical and adaptable way to incorporate trustworthiness considerations into AI design, development, deployment, use, and evaluation. citeturn0search0turn0search11
The framework was developed through an open, transparent, multidisciplinary and multistakeholder process involving hundreds of organizations and public feedback. citeturn0search11
The four functions of NIST AI RMF
The AI RMF Core contains four functions:
GOVERN → MAP → MEASURE → MANAGE
They should not be interpreted as a simple linear checklist. NIST describes Govern as cross-cutting and the overall risk management process as continuous throughout the AI system lifecycle. citeturn0search5turn0search29
Govern
Govern establishes organizational structures and practices for AI risk management.
It addresses areas such as policies, accountability, roles and responsibilities, organizational culture, legal and regulatory awareness, risk tolerance, and processes for integrating trustworthy-AI considerations.
In GRC terms, Govern is where AI risk becomes an organizational responsibility rather than an isolated technical activity.
Typical evidence may include:
AI governance policy;
assigned accountable owners;
approval and escalation criteria;
risk appetite or tolerance statements;
roles and responsibilities;
training requirements;
governance committee records;
policy exceptions and decisions.
Map
Map establishes the context in which AI operates and identifies risks associated with that context.
The organization should understand the intended purpose, users, affected groups, operating environment, assumptions, limitations, dependencies, benefits, potential harms, and foreseeable misuse.
NIST explains that Map outputs provide the basis for the Measure and Manage functions and can inform an initial decision about whether an AI system should proceed. citeturn0search34
For GRC, useful Map artifacts include:
AI use-case records;
AI system inventory;
stakeholder analysis;
impact assessments;
data and dependency mapping;
risk scenarios;
intended-use documentation;
limitation and assumption registers.
Measure
Measure applies appropriate methods and metrics to assess, analyze, and track AI risks.
NIST emphasizes that measurement should begin with the most significant identified risks and that risks or trustworthiness characteristics that cannot be meaningfully measured should be documented. citeturn0search14turn0search5
Depending on the system, measurement may address:
performance and accuracy;
validity and reliability;
robustness;
security and resilience;
privacy;
fairness;
transparency;
explainability and interpretability;
human oversight;
potential impacts.
Measurement is not a one-time activity. NIST states that methods, contexts, risks, and impacts can evolve, requiring continued application of Measure. citeturn0search5
Manage
Manage uses outputs from Map and Measure to prioritize and respond to AI risks.
NIST's Core includes determining whether an AI system achieves its intended purpose and whether development or deployment should proceed. Risk treatment is then prioritized based on factors such as impact, likelihood, and available resources or methods. citeturn0search13turn0search5
Manage may result in decisions to:
mitigate risk;
modify a system;
restrict deployment;
introduce additional controls;
increase human oversight;
accept residual risk within approved tolerance;
or determine that deployment should not proceed.
The important point is that AI risk management should lead to decisions, not merely documentation.
NIST AI RMF Core: categories and subcategories
Each AI RMF function is divided into categories and subcategories.
The structure gives organizations progressively more specific outcomes and actions without requiring every organization to implement them identically.
This design supports proportional implementation: a low-risk internal use case does not necessarily require the same governance, measurement, and evidence as a high-impact AI system.
NIST explicitly states that the actions in the Core do not constitute a checklist and are not necessarily an ordered set of steps. citeturn0search5
NIST AI RMF and the AI lifecycle
AI risk management should continue across the lifecycle rather than stop at deployment.
A practical lifecycle view is:
Plan → Design → Develop → Test → Deploy → Operate → Monitor → Change or Retire
At each stage, the organization can revisit context, risks, controls, measurements, and decisions.
This matters because an AI system can change after deployment through:
new data;
model updates;
changed prompts or configurations;
new users;
vendor changes;
new integrations;
changing threats;
changing regulations;
changes in the business context.
The AI RMF is designed to adapt to these evolving conditions. citeturn0search0turn0search5
NIST AI RMF and trustworthy AI
NIST's framework connects risk management with characteristics of trustworthy AI.
Depending on context, organizations may need to consider characteristics such as:
valid and reliable;
safe;
secure and resilient;
accountable and transparent;
explainable and interpretable;
privacy-enhanced;
fair with harmful bias managed.
The relevant characteristics and their relative importance can differ by use case. The purpose is not to maximize every characteristic equally, but to understand the tradeoffs and risks associated with the system.
Is NIST AI RMF mandatory?
NIST AI RMF 1.0 is intended for voluntary use. NIST describes it as a voluntary resource for organizations designing, developing, deploying, or using AI systems. citeturn0search0turn0search3
However, an organization may still have mandatory obligations from laws, regulations, contracts, sector requirements, or internal policies. Those requirements should be assessed separately and then connected to the AI risk management process where appropriate.
Is NIST AI RMF a certification?
The NIST AI RMF itself is not presented by NIST as a certification scheme.
Organizations can use the framework to structure internal governance, risk management, assessments, controls, evidence, and reporting. Whether an organization needs an external assessment or certification depends on other standards, contractual requirements, regulatory expectations, or assurance objectives.
NIST AI RMF Playbook
The NIST AI RMF Playbook is a companion resource for voluntary use. It provides suggested actions, references, and related guidance intended to help organizations achieve outcomes associated with Govern, Map, Measure, and Manage. citeturn0search2
NIST states that organizations may use as many or as few suggestions as apply to their industry, use case, or interests.
The distinction is important:
AI RMF = framework
Playbook = implementation guidance and suggested actions
The Playbook is not a mandatory checklist, and NIST says it will be updated after AI RMF 1.0 is revised. citeturn0search2
NIST AI RMF and GRC
For a GRC organization, NIST AI RMF can be mapped into existing governance and risk processes.
The AI RMF functions can then align with existing GRC capabilities:
NIST AI RMF
GRC capability
Govern
Policies, accountability, compliance
Map
Risk assessment, impact analysis
Measure
Testing, evaluation, evidence
Manage
Treatment, issues, remediation
Continuous monitoring
KRIs, dashboards, management reporting
This avoids creating a completely separate AI risk program when existing enterprise risk and control infrastructure can support it.
NIST AI RMF crosswalks
NIST maintains crosswalk resources that map AI RMF 1.0 to other frameworks and standards. NIST's crosswalk page notes that these resources are intended to help organizations use AI RMF and align it with related guidance. citeturn0search1
The public NIST ecosystem also includes community-submitted crosswalks. NIST makes clear that inclusion of a community crosswalk does not constitute NIST endorsement of the mapped resource. citeturn0search8
For GRC teams, crosswalks can reduce duplicate control mapping and help identify where existing governance capabilities already support AI risk requirements.
NIST AI RMF and Generative AI
NIST published AI 600-1, Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, on July 26, 2024.
The profile is a cross-sectoral companion resource to AI RMF 1.0 focused on generative AI. It helps organizations identify and manage risks associated with generative AI while using the broader AI RMF structure. citeturn0search17
This means organizations do not need to abandon the AI RMF Core when dealing with GenAI. Instead, the Generative AI Profile adds context and risk-management considerations for that technology.
NIST AI RMF implementation example
Consider an organization introducing an AI system into a customer-facing process.
Govern
The organization defines ownership, acceptable use, risk tolerance, approval criteria, and escalation requirements.
Map
The team documents the purpose, users, affected customers, data, dependencies, expected benefits, foreseeable harms, and operating context.
Measure
The organization establishes appropriate testing and evaluation methods for performance, security, privacy, fairness, reliability, or other material characteristics.
Manage
The organization decides whether the system can proceed, what controls are required, which risks require treatment, and who owns the remaining risk.
Monitor
The organization continues to evaluate the system as conditions change and escalates material issues.
The exact controls and metrics depend on the use case. NIST AI RMF provides the structure; it does not prescribe one universal control set.
What evidence should an AI RMF program maintain?
A mature program should be able to demonstrate how decisions were made.
Potential evidence includes:
AI system inventory;
use-case description;
accountable owner;
risk assessment;
impact analysis;
risk scenarios;
control mappings;
testing and evaluation results;
approval records;
vendor assessments;
incidents and issues;
remediation;
residual-risk acceptance;
monitoring results;
management reporting.
This evidence creates traceability from governance decisions to operational outcomes.
How to measure AI RMF maturity
A practical GRC maturity model can use five levels:
Level 1 — Ad hoc
AI use exists with limited consistent governance or documentation.
Level 2 — Defined
Policies, ownership, inventory, and basic risk assessment practices exist.
Level 3 — Operational
Govern, Map, Measure, and Manage are implemented through repeatable workflows.
Level 4 — Integrated
AI risk is integrated with ERM, compliance, security, privacy, third-party risk, and internal audit.
Level 5 — Continuous
Monitoring, measurement, lessons learned, and control improvements operate continuously.
This is a practical GRC interpretation, not an official NIST maturity model.
Common mistakes when implementing NIST AI RMF
Treating the framework as a checklist
NIST explicitly states that AI RMF actions are not a checklist or necessarily an ordered sequence. citeturn0search5
Starting with controls instead of context
Map exists for a reason: organizations need to understand the system and its context before deciding what risks and controls matter.
Measuring everything equally
Measurement should prioritize significant risks and use appropriate methods.
Treating deployment as the end
Risk management continues throughout the lifecycle.
Creating a parallel GRC process
Where possible, integrate AI risk into existing enterprise risk, compliance, security, privacy, vendor-risk, issue-management, and audit capabilities.
Ignoring framework updates
NIST currently states that AI RMF 1.0 is being revised. Organizations should monitor the revision process and review their implementation when the revised framework becomes available. citeturn0search6
Frequently asked questions
What is NIST AI RMF 1.0?
It is NIST's voluntary Artificial Intelligence Risk Management Framework, published January 26, 2023, to help organizations manage AI risks and promote trustworthy and responsible AI. citeturn0search0
What are the four NIST AI RMF functions?
Govern, Map, Measure, and Manage.
What is the NIST AI RMF Core?
It is the part of AI RMF that organizes AI risk management into four functions, with categories, subcategories, outcomes, and actions. citeturn0search5
Is NIST AI RMF a standard?
It is a NIST framework and voluntary resource. It can be used alongside standards and other governance frameworks.
Is NIST AI RMF mandatory?
AI RMF 1.0 itself is voluntary. Other legal, regulatory, contractual, or organizational requirements may still apply.
What is the NIST AI RMF Playbook?
It is a companion resource containing suggested actions and references for using the AI RMF.
What is the difference between NIST AI RMF and the AI RMF Playbook?
The AI RMF provides the framework and Core; the Playbook provides suggested implementation actions and references.
Is NIST AI RMF being updated?
Yes. NIST's current AI RMF website states that AI RMF 1.0 is being revised. citeturn0search6
Does NIST AI RMF cover generative AI?
Yes. NIST published a Generative AI Profile in 2024 as a companion resource to AI RMF 1.0. citeturn0search17
Conclusion
The NIST AI Risk Management Framework is best understood as a flexible operating structure for managing AI risk, not as a checklist or a single compliance target.
Its four functions — Govern, Map, Measure, and Manage — create a common language for governance, contextual risk identification, evaluation, treatment, and ongoing monitoring.
For GRC teams, the strongest implementation strategy is usually to connect AI RMF with existing enterprise capabilities: risk management, controls, compliance, security, privacy, third-party risk, issue management, audit, and executive reporting.
The framework remains highly relevant as organizations move from isolated AI experimentation toward enterprise-scale governance. At the same time, NIST currently states that AI RMF 1.0 is being revised, making version awareness an important part of any long-term implementation strategy. citeturn0search6
