The civic record

Read. Follow. Verify.

HR8283FEDERALSession 119Federal Billin committee

High Impact

Deterring American AI Model Theft Act of 2026

Last action · April 22, 2026: Ordered to be Reported (Amended) by the Yeas and Nays: 43 - 0.Sponsors: Bill Huizenga and 21 others

Bill progressHistory & official text ↓

Introduced

Committee

Floor Vote

Passed

Law

Track this bill to get notified when it advances a stage. One tap to stop, anytime.

What This Bill Would Do

This bill aims to protect American-owned, closed-source artificial intelligence models from unauthorized data extraction by foreign adversaries. It establishes legal definitions for 's' and 'entities of concern' to help the U.S. government identify and deter the theft of proprietary AI technical features.

Why it matters

The bill creates a federal framework to identify and potentially penalize foreign entities that bypass security controls to steal proprietary AI technology, impacting how U.S. AI companies manage access to their models.

Who It Would Affect

Groups the bill's text would affect if enacted, as described in the record. Being affected is not a position for or against it.

Directly affected

U.S. AI Companies

These companies are the owners of closed-source models and would be the primary entities protected by the identification and deterrence of extraction attacks.

Directly affected

Foreign entities in countries of concern

These entities face potential identification and penalties if they are found to be conducting unauthorized model extraction attacks.

Key Facts

Defines 'closed-source AI models' as proprietary models with restricted access, governed by terms of service, and not publicly available.
Defines 'model extraction attacks' as unauthorized querying to replicate or train other AI models by circumventing access controls or violating terms of service.
Identifies 'countries of concern' as China (including Hong Kong and Macau), Russia, and other nations designated by the Secretary of State.
Defines 'entities of concern' as foreign persons or entities headquartered in countries of concern or those conducting unauthorized model extraction.

Show all 9 facts

Defines 'fraudulent account network providers' as foreign entities that knowingly provide access to restricted AI models for entities of concern.
Excludes from the definition of 'fraudulent account network provider' entities that provide internet access for freedom of expression purposes.
Allows the purpose of a model extraction attack to be inferred from query volume, patterns, coordination, and correlation with other AI development timelines.
States that authorized model training consistent with terms of service is a legitimate research method and distinct from extraction attacks.
Designates the House Committee on Foreign Affairs and the Senate Committee on Banking, Housing, and Urban Affairs as the appropriate congressional committees for oversight.

Frequently Asked Questions

Does this bill stop me from using AI for research?

No. The bill explicitly states that authorized model training that follows the owner's terms of service is a legitimate research method and is not considered a .

What counts as a 'model extraction attack'?

It is the unauthorized extraction of an AI model's capabilities to replicate or improve another model by bypassing security controls, using fraudulent credentials, or violating the owner's terms of service.

What else do you want to know?

News Coverage

No news coverage found yet. Articles are indexed twice daily.

Follow the record

Votes, people, money & the original source

Where This Stands

Currently in committee. The next step in the legislative lifecycle is Floor Vote.

Procedural history

4 actions

Apr 22, 2026HouseCommittee
Ordered to be Reported (Amended) by the Yeas and Nays: 43 - 0.
Apr 22, 2026House
Committee Consideration and Mark-up Session Held
Apr 15, 2026HouseCommittee
Referred to the House Committee on Foreign Affairs.
Apr 15, 2026HouseIntroduced
Introduced in House

Version history

Only one version on file - nothing to compare yet. As later stages (committee substitute, engrossed, enrolled) are captured, the redline appears here.

View official text →

Full breakdown

Who this bill affects and the strongest case for and against it, written only from the bill’s own text.

Discoveries

1 pattern noticed across the record

Patterns POLISCOPE noticed across the record. These are observations to investigate, not conclusions.

policy shift90% confidence

Inference-based Enforcement

The bill allows the government to infer malicious intent based on the 'totality of circumstances' of query patterns, rather than requiring direct evidence of intent.

United States Government

Connected Entities

Huizenga, Mr., Russian Federation, John Moolenaar and 1 more

PersonHuizenga, Mr.“Primary sponsor of the bill in the House of Representatives.”Map →

locationRussian Federation“Explicitly named as a country of concern.”Map →

PersonJohn Moolenaar“Co-sponsor of the bill.”Map →

locationPeople's Republic of China“Explicitly named as a country of concern.”Map →

Sources

Open source document

www.congress.gov

Analysis Score

0–100

Significance85
How much this matters to a regular citizen
Controversy40
Intensity of disagreement among stakeholders
Entertainment10
Compellingness for a non-policy-wonk reader
Buzz60
Current news / social attention level

Publisher tools