AI Kill Switch Act would let Trump admin order shutdown of rogue AI systems
Ars Technica Jon Brodkin ● Covered by 3 sources
A new bill would let the government order AI companies to shut down systems it deems dangerous, on demand. Built-in kill switches, huge daily fines for refusal — this is real regulatory teeth, not a guideline.
Based on reporting by Ars Technica, Jon Brodkin — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
A bipartisan group in Congress has introduced something with a blunt, almost cinematic name: the AI Kill Switch Act. Strip away the branding and what's left is a genuine expansion of federal power over private AI infrastructure. The bill would amend the Homeland Security Act of 2002 to give the Secretary of Homeland Security explicit authority to order AI companies to suspend or fully shut down systems judged capable of causing catastrophic harm.
This isn't a polite request. Under the proposal, AI developers would be legally required to build in technical mechanisms — throttles, access blocks, capability restrictions, or a full kill switch — so the government can act the moment it decides a system poses a serious risk. Refuse to comply and the penalties are steep: up to $20 million per day the violation continues. That's not a fine you shrug off; it's a number designed to make noncompliance financially irrational even for a company flush with venture capital.
What's notable is who would hold this authority. The bill hands the power to the sitting administration, meaning the Trump administration right now, and whoever comes after. Given how contested the definition of
My take — AI-written commentary, not fact-checked reporting
I'm generally skeptical of Washington's ability to regulate anything moving this fast, but a government-mandated kill switch is the kind of blunt instrument that sounds reassuring until you remember who gets to press it — and that concentration of shutdown power in a single administration's hands is its own catastrophic-harm scenario, especially compared to how the EU has tried, however clumsily, to build risk tiers instead of an off-switch.
Read more about this at: Ars Technica
