Overview
Severity: MEDIUM | Affected: NIST | Category: policy
The U.S. National Institute of Standards and Technology (NIST) has released the final version of its AI Secure Development Framework (AI-SDF), a set of standards and best practices for building secure and trustworthy AI systems. A key mandate of the new framework is the requirement for a comprehensive AI Bill of Materials (AIBOM), which must track the provenance of all training data, model components, and dependencies. This is intended to increase transparency and help organizations identify risks from poisoned data or vulnerable pre-trained models. The framework will be mandatory for all AI systems procured by the U.S. federal government starting in Q3 2026. Industry experts view the AI-SDF as a landmark policy that will push the private sector towards adopting more rigorous security practices throughout the AI/ML lifecycle.