Deterring American AI Model Theft Act of 2026
Sponsored By: Representative Huizenga, Bill [R-MI-4]
In Committee
Summary
This bill would create a cross-agency framework to deter and punish the unauthorized extraction of closed-source U.S. AI models. It pairs coordinated assessments, a public attackers list, and export-control and emergency-sanction tools to identify and block model theft.
Show full summary
- Model owners and AI developers would get a formal, confidential info-sharing channel with the Department of Commerce, published best-practice guidance, and a role in public consultations during an initial assessment due in 180 days and a Congress report due in 210 days.
- Federal agencies would gain a structured enforcement path that uses existing Export Control Reform Act and International Emergency Economic Powers Act authorities and committee reviews to designate and sanction attackers.
- Foreign actors and providers of fraudulent account networks could be added to a State Department "AI Model Extraction Attackers List" visible for up to five years and face Entity List restrictions and blocked transactions under IEEPA.
Your PRIA Score
Personalized for You
How does this bill affect your finances?
Sign up for a PRIA Policy Scan to see your personalized alignment score for this bill and every other piece of legislation we track. We analyze your financial profile against policy provisions to show you exactly what matters to your wallet.
Bill Overview
Analyzed Economic Effects
3 provisions identified: 2 benefits, 0 costs, 1 mixed.
New rules to protect model owners
If enacted, the State and Commerce Departments would complete an assessment within 180 days naming who has run or helped model extraction attacks and which countries they came from. Commerce would hold voluntary public consultations with owners and experts and give a report to Congress within 210 days. Commerce would also give annual updates for three years and keep monitoring attacks after that. The bill would set definitions for closed‑source models and countries of concern like China and Russia to decide who is covered.
Public list of AI attackers
If enacted, the State Department would keep a public "AI Model Extraction Attackers List" of people and companies found to have done or directed extraction attacks in the past year. The list could be posted publicly for up to five years. The State Department would not disclose confidential information that owners gave without the owners' permission.
Sanctions and export limits for attackers
If enacted, Commerce would decide within 210 days whether to add named attackers and their affiliates (50%+ ownership) to the Entity List, which limits exports and sales to those entities. The President, acting through State, would also be authorized to block or freeze property and transactions under the International Emergency Economic Powers Act for designated actors. There would be exceptions for U.N. obligations, humanitarian aid, and authorized U.S. intelligence, law enforcement, and national security activities. Violating these blocking orders would carry civil and criminal penalties under IEEPA.
Sponsors & CoSponsors
Sponsor
Huizenga, Bill [R-MI-4]
MI • R
Cosponsors
Rep. Moolenaar, John R. [R-MI-2]
MI • R
Sponsored 4/15/2026
Rep. Lawler, Michael [R-NY-17]
NY • R
Sponsored 4/20/2026
Crawford
AR • R
Sponsored 4/21/2026
Rep. Kamlager-Dove, Sydney [D-CA-37]
CA • D
Sponsored 4/29/2026
LaHood
IL • R
Sponsored 4/29/2026
Rep. Cline, Ben [R-VA-6]
VA • R
Sponsored 4/29/2026
Rep. Harrigan, Pat [R-NC-10]
NC • R
Sponsored 4/29/2026
Rep. McCaul, Michael T. [R-TX-10]
TX • R
Sponsored 5/11/2026
Rep. Moore, Barry [R-AL-1]
AL • R
Sponsored 5/11/2026
Salazar
FL • R
Sponsored 5/12/2026
Rep. James, John [R-MI-10]
MI • R
Sponsored 5/12/2026
Roll Call Votes
No roll call votes available for this bill.
View on Congress.gov