Bill requires big AI firms to give models to federal labs.
H.R. 9965 — AI Threat Output and Monitoring Incident Containment Act · Filed by Celeste Maloy (R-UT) · 2 cosponsors · Introduced Jul 27, 2026 · Referred to committee
Your members of Congress
Enter a ZIP to see where your representative and both senators stood on this bill.
Looked up on this device — your ZIP is never stored on our servers.
What it does
The Department of Energy must set up a program to test AI systems. The tests check if AI could be used as weapons or cause nuclear harm. Large AI firms that spend $2 billion or more on AI must join. They must give their AI models to the agency to test. Tests will try to break the AI's safety rules. The agency can fine firms up to $1 million per day if they refuse.
Who it affects
Large AI firms must share their AI models with the agency. People could be safer if tests find ways to stop AI from causing harm. The agency gets new power to test AI and fine firms.
One thing to notice
Only about 3 to 5 firms worldwide would have to follow this rule. Smaller AI firms and open-source AI projects would not have to join.
From the analysis of the bill text, linked under Primary records below.
Where it stands
2 cosponsors: 1 Democrats, 1 Republicans.
- Jul 27, 2026 — Introduced · Congress.gov: “Introduced in House”
- Jul 27, 2026 — Referred to House Committee on Science, Space, and Technology · Congress.gov: “Referred to the House Committee on Science, Space, and Technology”
Dates and quoted wording are Congress.gov's action record; the timeline shows status changes, not every procedural step.
Words to know
- Department of Energy — The federal agency that handles energy and nuclear safety for the nation.
- AI models — The trained AI systems that firms build to perform tasks and make decisions.
- safety rules — Built-in limits that stop an AI from doing harmful or dangerous things.
How this was measured
Analysis — Quorum's AI read the bill text published by Congress.gov (14,413 characters) on Jul 31, 2026. Section numbers in the findings refer to that text, linked below; transparency and hidden-provision scores are compared against the median of 14,206 analysed bills.
Status and sponsors — Congress.gov's bill record — actions, committee referrals and cosponsors — loaded nightly. The timeline shows status changes, not every procedural action.
As of — page rendered 2026-09-17.
“Bill requires big AI firms to give models to federal labs.” QuorumCivic. https://share.quorumcivic.app/bill/119/hr9965/simple Report an error