The NIST AI Risk Management Framework (AI RMF) is a voluntary framework from the US National Institute of Standards and Technology for managing the risks of AI systems, organised into four functions: Govern, Map, Measure and Manage.
What AI RMF means
NIST released AI RMF 1.0, published as NIST AI 100-1, on 26 January 2023. It describes the characteristics of trustworthy AI: valid and reliable, safe, secure and resilient, accountable and transparent, explainable and interpretable, privacy-enhanced, and fair with harmful bias managed. A companion Playbook suggests actions for each function.
In July 2024 NIST published the Generative AI Profile, NIST AI 600-1, which applies the framework to risks that are specific to or made worse by generative AI, such as confabulation, data privacy and information security. The Govern function includes outcomes on risks from third-party software, data and other supply chain issues.
Vendors in the US often describe their AI governance against the AI RMF. Ask for the mapping, but verify it against concrete evidence such as testing results and model documentation. The AI RMF can be mapped to ISO/IEC 42001 or the EU AI Act, but each has its own requirements.
What to check in vendor documents
- Whether the vendor maps its AI governance to the AI RMF functions.
- Evidence of testing for validity, reliability and bias in your use case.
- How the vendor manages third-party models and data, as the Govern function expects.
- Use of the Generative AI Profile for generative features.
Related terms
- ISO/IEC 42001 (AIMS)ISO/IEC 42001 is the international standard that specifies requirements for an artificial intelligence management system (AIMS), covering how an organisation that develops, provides or uses AI governs it responsibly. Organisations can be certified against it.
- EU AI ActThe EU AI Act is Regulation (EU) 2024/1689, the EU's risk-based law on artificial intelligence. It bans some practices, sets duties for high-risk systems, general-purpose AI models and transparency, and splits obligations between providers and deployers.
- Explainable AI findingAn explainable AI finding is a result produced with AI assistance that shows its basis: the rule it was checked against, the evidence excerpt and location it relies on, and enough context for a human reviewer to verify or reject it.
- AI vendor riskAI vendor risk is the risk an organisation faces when it buys a product that relies on AI models: how its data is used for training, which model providers process it, how outputs are controlled and how the vendor meets AI rules.
- NIST Cybersecurity Framework 2.0 (NIST CSF 2.0)NIST CSF 2.0 is the second version of the US National Institute of Standards and Technology's Cybersecurity Framework, published on 26 February 2024. It organises outcomes into six functions, with supply chain risk management under the new Govern function.
Sources
- NIST, AI Risk Management Frameworkwww.nist.gov/itl/ai-risk-management-framework
- NIST AI 100-1, Artificial Intelligence Risk Management Framework (AI RMF 1.0)doi.org/10.6028/NIST.AI.100-1
- NIST AI 600-1, Generative Artificial Intelligence Profiledoi.org/10.6028/NIST.AI.600-1
Summarised for practitioners, not legal advice. Check the source text for your own obligations.