Sponsors (21)
Official title: To prevent foreign adversaries from threatening the national security of the United States by extracting key technical features of closed-source, American-owned artificial intelligence models, and for other purposes.
Introduced April 15, 2026 by Bill Huizenga · Last progress April 15, 2026
The bill strengthens U.S. defenses and legal tools to protect proprietary AI models and block malicious foreign acquisition, at the trade‑off of greater executive control, compliance burdens for firms, and risks of chilling legitimate research and diplomatic fallout.
U.S. model owners and the broader tech sector gain coordinated federal support to detect, deter, and respond to model-extraction attacks, strengthening protection of proprietary AI capabilities and U.S. competitiveness.
U.S. companies get clearer legal definitions (closed‑source models, model extraction) and clarified export/reexport/in‑country transfer terms, making it easier to enforce IP rights and comply with export controls.
The bill creates targeted export-control tools and a time‑bound interagency (210‑day) process, plus listing/blocking authorities, to identify and restrict malicious actors that steal or misuse U.S. AI models and components.
Researchers, developers, and academia could be chilled from legitimate research, defensive interoperability testing, or international collaboration because broad definitions and enforcement risks are unclear or overbroad.
Smaller companies and some financial firms will face increased compliance, reporting, and potential law‑enforcement involvement, raising costs and diverting resources from product development.
Concentrated executive designation and blocking powers (Entity List / IEEPA authorities) create opaque decisionmaking risk and may lead to broad economic or foreign‑policy impacts without clear congressional oversight.
Based on analysis of 5 sections of legislative text.
Requires interagency assessments of foreign model extraction attackers, a public attackers list, best‑practice guidance, Entity List additions, and IEEPA sanctions authority against identified entities.
Creates a federal process to identify foreign actors who steal or shortcut access to U.S. closed‑source AI model capabilities through model extraction attacks and fraudulent account networks, and to deter them through export‑control listings and IEEPA sanctions. It requires interagency assessments, outreach to affected model owners, an owner‑protected public attackers list, best‑practice guidance, and a decision process to add offending entities to the Commerce Department Entity List. Sets timelines for initial assessments and reports (180–210 days), ongoing information sharing and updates for several years, and grants authority for the President to block property and transactions of identified entities under IEEPA, with limited humanitarian and intelligence exceptions.