National Defense Authorization Act for Fiscal Year 2026, Section 6603 ("Application of artificial intelligence policies of the intelligence community to publicly available models hosted in classified environments")

• Authority: United States Congress
• Parent authority: Federal government
• Jurisdiction: United States
• Collections: NDAA provisions;U.S. federal laws;FY2026 NDAA
• Most recent activity: Enacted
• Most recent activity date: 2025-12-18

Risk subdomain coverage

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• AI system security vulnerabilities
• Governance failure
• Lack of capability or robustness
• Lack of transparency or interpretability

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has minimal coverage of risk domains, with primary focus on AI system security (2.2) through testing standards and secure computing environments. There is implicit coverage of governance failure (6.5) through establishment of oversight mechanisms, and minimal coverage of system safety (7.3, 7.4) through testing requirements. The document is primarily procedural rather than risk-focused.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• National Security

All other sectors had no coverage in this document.

Sectors Summary

This document exclusively governs the National Security sector, specifically the intelligence community's use of AI models in classified environments. It establishes testing standards, secure computing requirements, and policy application for AI models used by intelligence agencies.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding federal statute enacted by the United States Congress as part of the National Defense Authorization Act. It contains mandatory language ('shall') establishing legal obligations for the Director of National Intelligence and intelligence community officers, with clear enforcement authority through the executive branch.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: United States Congress
• Summary: The document is Section 6603 of the National Defense Authorization Act for Fiscal Year 2026, which is federal legislation enacted by the United States Congress.

Target

• Entity Types: AI_Governance_Actor, AI_Deployer
• Entities: Director of National Intelligence; Chief Artificial Intelligence Officer of the Intelligence Community; elements of the intelligence community; vendors or prospective vendors
• Summary: The document applies to intelligence community entities and their officers who must establish policies and testing standards for AI models. It also indirectly affects vendors who provide AI models to the intelligence community, though it explicitly prohibits directing vendors to alter models for viewpoint bias.

Enforcer

• Entity Types: AI_Governance_Actor
• Entities: Director of National Intelligence; Chief Artificial Intelligence Officer of the Intelligence Community
• Summary: The Director of National Intelligence has authority to ensure policy application and make revisions, while the Chief AI Officer establishes and enforces testing standards across the intelligence community.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: Chief Artificial Intelligence Officer of the Intelligence Community; Chief Artificial Intelligence Officers of the elements of the intelligence community
• Summary: The Chief AI Officer coordinates with element-level AI officers to set standards for testing and evaluation, implying ongoing monitoring responsibilities for model performance and compliance.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document primarily covers the Verify and Validate stage through establishment of testing standards and benchmarks, and the Deploy stage through requirements for secure computing environments and pre-acquisition evaluation. It also addresses Operate and Monitor through policy application to models in classified environments.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• Open Weight / Open Source
• Task-specific AI

Technical Scope Summary

The document explicitly mentions AI models and focuses on publicly available models hosted in classified environments. It does not define specific AI types (frontier, general purpose, generative, etc.) or mention compute thresholds or open-weight models. The scope is broad, covering models 'generally available to the public' when used in intelligence community contexts.