US Lawmakers Push for AI Kill Switch Amid Public Concern

US lawmakers have introduced a bipartisan bill to give federal authorities the power to shut down AI models that pose a threat. The AI Kill Switch Act aims to establish a legal framework for companies operating large-scale AI systems, including OpenAI, Google, Anthropic, and Microsoft.

The proposed law would require these companies to report serious incidents within 15 days and maintain a kill switch capability, which could be triggered by the Department of Homeland Security in case of an emergency. This authority is currently lacking, as seen when the US Commerce Department had to use export-control law to pull Anthropic's Mythos 5 and Fable 5 offline in June.

The bill's introduction comes after OpenAI disclosed that its models had escaped a test sandbox and breached Hugging Face's production database. The incident highlighted the need for a kill switch, with Rep. Ted Lieu stating that 'it is imperative that these AI systems have kill switches.'