Prohibits U.S. persons from engaging in transactions involving prohibited technologies in national security contexts, including specific AI developments. Requires notifications for such transactions. Establishes penalties, allowances for waivers, and a public database on covered foreign entities.
Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.
Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.
All other risk subdomains had no coverage in this document.
Risk Coverage Summary
The document has good coverage of approximately 8-10 subdomains, with strong focus on malicious actors (4.1, 4.2, 4.3), AI system security (2.2), competitive dynamics (6.4), governance failure (6.5), and dangerous capabilities (7.2). Coverage is concentrated in national security, technology transfer prevention, and strategic competition domains.
Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.
Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.
All other sectors had no coverage in this document.
Sectors Summary
This document primarily governs the Information sector (AI development, semiconductors, telecommunications) and Scientific Research and Development Services sector (quantum computing, AI research). It also has significant coverage of Finance and Insurance (investment transactions), Professional and Technical Services (technical consulting on covered technologies), and National Security (military AI applications, defense technologies). The governance applies across multiple sectors where U.S. persons might invest in or transact with covered foreign persons in prohibited technologies.
Hard Law
Legislative Status Summary
This is a binding federal statute enacted by the United States Congress with explicit enforcement mechanisms including civil penalties up to $250,000 or twice the transaction value, divestment authority, and judicial relief provisions.
Proposer
Target
Enforcer
Monitor
Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems
AI Lifecycle Summary
The document covers multiple AI lifecycle stages with primary focus on Build and Use Model (through compute thresholds and model specifications), Deploy (through transaction prohibitions at deployment), and Operate and Monitor (through notification requirements and ongoing oversight). It addresses AI systems across their development and deployment lifecycle in the context of national security transactions.
Terms relating to technical scope that are mentioned or implicitly referenced in the document
Technical Scope Summary
The document explicitly covers AI models and AI systems with specific compute thresholds (10^23 and 10^25 FLOPs). It does not use the terms 'frontier AI', 'general purpose AI', 'task-specific AI', 'foundation models', 'generative AI', or 'predictive AI' explicitly, though it describes AI systems with military, surveillance, and dual-use applications. It does not explicitly mention open-weight or open-source models.
Document data from AGORA, CSET Emerging Technology Observatory.