WhoVotedWhy
119th Congress · HR.9917In committeeCongress.gov ↗JSONReport an error

AI Kill Switch Act

Summary, written by a model from the CRS summary

Would require AI systems to have a mechanism to be shut down or disabled under defined conditions. It aims to give operators or regulators a way to halt an AI system that behaves dangerously.

Model-rated AI relevance 100%.

safetytransparency
TL

Sponsored by

Ted LieuDemocratCA

Data as of: bills Sep 9, 2026 · votes Sep 11, 2026 · statements Sep 9, 2026. What changed

Progress

  1. Introduced

    Jul 23, 2026

  2. Reported by committee

  3. Passed House

  4. Passed Senate

  5. Became law

Latest action · July 24, 2026

Referred to the Subcommittee on Cybersecurity and Infrastructure Protection.

What members said

Releases that cite this bill first, then releases on the same topics.

Nathaniel MoranRepublicantoward oversightAug 19, 2026 · cites this bill by title

Moran champions the bipartisan AI Kill Switch Act requiring frontier AI developers to build human intervention and shutdown capabilities, framing it as light-touch safety that does not impede American AI leadership.

Source ↗
Ted LieuDemocrattoward oversightJul 23, 2026 · cites this bill by title

Lieu's AI Kill Switch Act would require large AI developers to be able to shut down powerful systems and give the government emergency shutdown authority.

Source ↗
Bernard SandersIndependenttoward oversightSep 3, 2026 · same topic

Sanders would permanently ban superintelligent AI and pause advanced AI development until a new federal regulator sets safety rules.

Source ↗
Raphael WarnockDemocrattoward oversightSep 3, 2026 · same topic

Warnock calls for a statewide moratorium on AI data center construction unless projects guarantee no utility cost hikes, no NDAs, no tax giveaways, and no pollution.

Source ↗
Joshua HawleyRepublicantoward oversightSep 3, 2026 · same topic

Hawley is investigating AI surveillance camera companies, arguing Congress never authorized these invasive AI-driven networks and that corporate self-policing is an inadequate safeguard.

Source ↗
Tom BarrettRepublicantoward oversightSep 1, 2026 · same topic

Barrett's bipartisan Doctors Not AI Act would bar AI from making medical-necessity insurance denials and require disclosure of AI use.

Source ↗
Gabe EvansRepublicanmixedSep 1, 2026 · same topic

Evans promotes American open-source AI leadership against China while adding commonsense consumer guardrails and mandated federal risk assessments of adversary models.

Source ↗
Greg LandsmanDemocrattoward oversightSep 1, 2026 · same topic

Landsman's Doctors Not AI Act would keep licensed clinicians, not algorithms, in charge of medical-necessity decisions with transparency for patients.

Source ↗

Bill data read 2026-09-09T02:09:45.522Z