Bipartisan Bill Proposes Mandatory Kill Switches for Most Powerful AI Models
A bipartisan group of US lawmakers has introduced legislation that would require developers of the most powerful artificial intelligence models to install emergency kill switches, enabling the government to intervene if the technology poses national security risks.
Under the proposed bill, the Department of Homeland Security could order the throttling or complete shutdown of AI systems deemed dangerous. Companies failing to comply with such directives would face substantial financial penalties—up to $20 million per day.
The legislation targets frontier AI models, which represent the most advanced systems currently in development. Supporters argue the measure provides a necessary safety mechanism as AI capabilities continue to advance rapidly, while critics have raised concerns about potential overreach and the impact on innovation. The bill reflects growing bipartisan interest in establishing federal oversight of AI development, joining other recent proposals aimed at managing the risks associated with cutting-edge AI systems.