NIST AI Risk Management Framework
What you'll learn: How the NIST AI RMF organizes AI risk management into four complementary functions that work together systematically.
The Four-Function Structure
The National Institute of Standards and Technology (NIST) AI Risk Management Framework structures AI governance around four core functions that form a continuous lifecycle:
1. Govern
Establishes organizational culture, policies, and accountability structures for AI risk management. Think of this as setting the rules of the game—defining roles, responsibilities, and risk tolerances before you build anything.
2. Map
Identifies and contextualizes the specific risks in your AI system. This means understanding your use case, stakeholders affected, regulatory requirements, and potential harms. You're essentially drawing the terrain map of your project's risk landscape.
3. Measure
Quantifies and assesses the risks you've mapped using metrics, testing, and evaluation. This function answers: "How severe are these risks?" You might measure fairness metrics, accuracy across subgroups, or adversarial robustness—putting numbers to your concerns.
4. Manage
Takes action to mitigate, transfer, avoid, or accept identified risks. Based on measurements, you implement controls, monitoring systems, and response plans. This is where governance decisions become operational reality.
The Cycle in Action
These aren't sequential steps but interconnected functions. Governance informs what you map and measure; measurements feed back into governance decisions; managing risks reveals new areas to map. It's like a continuous quality improvement loop tailored for AI's unique challenges.
Key Takeaway: The NIST AI RMF's four functions—Govern, Map, Measure, and Manage—provide a systematic, cyclical approach to identifying, assessing, and mitigating AI risks throughout a system's lifecycle, emphasizing that effective risk management is ongoing, not one-time.