The National Institute of Standards and Technology (NIST) is developing the AI Risk Management Framework (AI RMF), a voluntary standard designed to make artificial intelligence systems safer and more trustworthy. Released in August as a second draft, the goal is to finalize version 1.0 in January. This framework offers a structured process for organizations to identify and manage risks related to bias, security, and privacy throughout the AI lifecycle.
The standard defines key characteristics for trustworthy AI, including fairness, reliability, and accountability, supported by a companion playbook and ongoing research into bias mitigation. While NIST also advances quantum information science, this specific initiative focuses on AI to foster innovation while protecting human rights and privacy. The institute is working with international partners to ensure the framework aligns with global standards, encouraging context-specific evaluation to identify risks before deployment.
Source: https://www.nist.gov/speech-testimony/trustworthy-ai-managing-risks-artificial-intelligence
Keywords: NIST, AI Risk Management Framework, bias mitigation
