Bill · S. 2938· Senate(119-S-2938)
Artificial Intelligence Risk Evaluation Act of 2025
- Introduced
- 9/29/2025
- Last activity
- 9/30/2026 — Committee on Homeland Security and Governmental Affairs. Hearings held.
- Sponsored by
- Josh Hawley
- Cosponsors
- - 1 - 1
- Introduced
- In Committee
What this bill would do
This bill, known as the "Artificial Intelligence Risk Evaluation Act of 2025," requires the Secretary of Energy to create the Advanced Artificial Intelligence Evaluation Program. The program is designed to perform testing and evaluation on advanced AI systems to identify risks such as loss of control, weaponization, or threats to critical infrastructure. Developers of these systems must participate in the program and provide requested technical information, including code and training data. The Act also aims to provide Congress with the data needed for future regulatory oversight. This oversight includes studying the potential rise of artificial superintelligence and its risks to humanity. The program is set to expire after seven years unless Congress chooses to renew it.
Who it would affect, and how
Potential Positive Impacts for People
-
Safeguarding Citizens: The program aims to provide Congress with the data and insights necessary to safeguard American citizens through Federal oversight (Section 2(b)).
-
Risk Identification: The program will offer standardized and classified testing to systematically collect data on the likelihood of "adverse AI incidents," such as threats to critical infrastructure or the safety of the public (Section 5(b)(1)).
-
Developing Safety Protocols: The program is tasked with developing recommended containment protocols, contingency planning, and mitigation strategies to address identified risks (Section 5(b)(5)).
-
Informed Regulation: The initiative seeks to inform the creation of evidence-based standards, regulatory options, and governance mechanisms (Section 5(b)(6)).
Potential Negative Impacts for People
-
High Financial Penalties: Developers who fail to participate in the program or violate deployment prohibitions face a fine of not less than $1,000,000 per day (Section 4(c)).
-
Strict Requirements for Developers: Companies must provide highly sensitive information to the Secretary, which can include underlying code, training data, and model weights (Section 4(a)(2)).
-
Restrictions on Commerce: No person is permitted to deploy an advanced artificial intelligence system for use in interstate or foreign commerce unless they are in compliance with the program's requirements (Section 4(b)).
-
Potential for Government Intervention: The program is tasked with developing options for regulatory or governmental oversight that could include potential nationalization of certain technologies (Section 5(b)(8)).
Groups and Interests Ranking
Every group the model scored this bill for, from -10 to +10. Hover a score for the reason it gave.
No significant interests or groups are affected by this bill.
- 9/30/2026
- Committee on Homeland Security and Governmental Affairs. Hearings held.
- 3/19/2026
- 1 Cosponsor added.
- 9/29/2025
- Read twice and referred to the Committee on Commerce, Science, and Transportation.
- 1 Cosponsor added.
- Introduced in Senate
Marsha BlackburnRepublican · Tennessee
Richard BlumenthalDemocrat · Connecticut