Voluntary framework to manage risks of artificial intelligence systems throughout their lifecycle, emphasizing trustworthiness, fairness, and transparency.

Published in January 2023, the NIST AI RMF provides a structured approach to identify, assess, and mitigate risks unique to AI systems including bias, lack of explainability, and safety failures. It organizes activities into four functions: Govern, Map, Measure, and Manage. Organizations use it to establish responsible AI practices before deployment. For example, a healthcare provider might apply the framework to validate that a diagnostic AI model produces equitable outcomes across demographic groups.