Celeste Maloy (R-UT)
AI Threat Output and Monitoring Incident Containment Act
A Republican bill with 1 Democratic cosponsor · See all 2 cosponsors
Last activity — Referred to the House Committee on Science, Space, and Technology.
In committee
Introduced and handed to a committee in the chamber it started in.
The committee holds it, and most bills are never heard from again. A hearing or a committee vote would be news; silence is the usual outcome, and it kills the bill without anyone voting against it.
Its path through Congress
- Introduced
- In Committee
What this bill would do
This bill directs the Secretary of Energy to establish the Advanced Artificial Intelligence Nuclear Evaluation Program through the National Laboratories. The program is designed to test and evaluate artificial intelligence to assess the likelihood of an AI nuclear incident. Large advanced artificial intelligence developers are required to participate in this program. The program includes testing for jailbreaking techniques, red-teaming, and facilitating third-party assessments. It also mandates the development of containment protocols and mitigation strategies for AI nuclear incidents or loss-of-control scenarios. Additionally, the Secretary must provide recommendations for future legislation based on the data collected.
Yes· 3 of 5
- Scored +8
Does this better protect the country from computer attacks?
The bill strengthens security by testing AI for vulnerabilities that could lead to nuclear incidents or loss-of-control scenarios.
- Scored +6
Does this put more money into scientific research?
The bill leverages National Laboratories to conduct specialized research into AI safety and nuclear risks.
- Scored +5
Does this better protect air, water and wildlife?
The program protects nuclear infrastructure and materials from being compromised by artificial intelligence.
The opposite
- Scored -4
Would business owners be better off under this?
Large AI developers face significant new regulatory burdens, potential civil penalties of $1 million, and subpoena powers.
- Scored -1
Does this make it easier to run a small business?
The bill targets large advanced AI developers rather than small businesses, which may create a regulatory environment focused on industry leaders.
Who it would affect, and how
Potential Positive Impacts for People
-
Risk Assessment and Safety: The program aims to test and evaluate artificial intelligence specifically to assess the likelihood of an AI nuclear incident (Section 2(a)(1)). This could help prevent dangerous outcomes involving nuclear weapons or materials.
-
Improved Security Protocols: The program requires the development of recommendations for containment protocols, contingency planning, and mitigation strategies regarding AI nuclear incidents and loss-of-control scenarios (Section 2(a)(5)).
-
Expert Testing: The bill incorporates testing protocols for red-teaming by entities with expertise comparable to sophisticated malicious actors to address practical jailbreaking techniques (Section 2(a)(2)).
-
Information Sharing for Participants: Each participant in the program receives a report detailing the outcomes of testing and evaluation, including the identification of risks and safety measures tested (Section 2(a)(4)).
-
Future Legislative Planning: The Secretary is required to assess data to identify trends and risks—such as scheming behavior—to develop recommendations for legislation to prevent such incidents (Section 2(f)(1)).
Potential Negative Impacts for People
-
Financial Penalties for Developers: Large advanced artificial intelligence developers may face civil penalties of up to $1 million for violating the provisions of this section, with each day of a continuing violation counting as a separate offense (Section 2(d)(1)(A)).
-
Compelled Production of Information: The Secretary has the authority to issue subpoenas to compel developers to produce artificial intelligence, including model weights and related software and documentation (Section 2(d)(2)).
-
Legal Risks: The Secretary may refer matters to the Attorney General to bring civil actions in district courts if a violation is believed to have occurred (Section 2(d)(3)).
-
Potential Disclosure of Information: While much information is exempt from disclosure, the Secretary may disclose participant information if it is required by a court order, requested by a Congressional committee, or determined to be necessary for national security, public safety, or critical infrastructure (Section 2(e)(2)).
Groups and Interests Ranking
Every group the model scored this bill for, from -10 to +10. Tap or hover a score for the reason it gave.
- National Security And Foreign Policy
- Does this better protect the country from computer attacks? — The bill strengthens security by testing AI for vulnerabilities that could lead to nuclear incidents or loss-of-control scenarios.
- Education And Technology
- Does this put more money into scientific research? — The bill leverages National Laboratories to conduct specialized research into AI safety and nuclear risks.
- Environment Energy And Infrastructure
- Does this better protect air, water and wildlife? — The program protects nuclear infrastructure and materials from being compromised by artificial intelligence.
- Demographic And Interest Groups
- Would business owners be better off under this? — Large AI developers face significant new regulatory burdens, potential civil penalties of $1 million, and subpoena powers.
- Governance And Ideology
- Does this make what the government does easier for the public to see? — The Secretary must submit regular reports to Congress regarding the findings of the AI evaluation program.
- Economy And Finance
- Does this make it easier to run a small business? — The bill targets large advanced AI developers rather than small businesses, which may create a regulatory environment focused on industry leaders.
- Social Issues And Justice
- Is this likely to lift people out of poverty? — Preventing catastrophic nuclear or AI-driven damage may indirectly protect economic stability for low-income populations.
- 8/20/2026
- 1 Cosponsor added.
- 7/27/2026
- Referred to the House Committee on Science, Space, and Technology.
- 1 Cosponsor added.
- Introduced in House
Tom BarrettRepublican · Michigan
Sara JacobsDemocrat · California