Lawmakers are preparing legislation that would create an “AI Kill Switch” framework and expand the Department of Homeland Security’s authority over certain AI systems. Multiple reports say the bill, expected to be introduced in the House on Thursday, is sponsored by Rep. Ted Lieu (D-California) and Rep. Nathaniel Moran (R-Texas). Under the proposal, AI developers would be required to maintain the technical capability to shut down or throttle their AI systems when directed. The bill would also authorize DHS to issue such orders, prompting companies to disable or limit their systems. The approach is described as requiring preparedness on the part of AI companies—ensuring they can comply with government instructions—while setting DHS as the agency able to trigger those actions. The reports do not specify additional details such as how DHS would define “rogue” or qualifying systems, how quickly companies must comply, or the legal process for issuing orders. The measure is characterized as a way to enable rapid interruption of AI behavior in response to DHS determinations, but it is still in the early stages of introduction and review.