[U.S. Congress Considers Including AI Kill Switch in Legal Requirements] DeepTech TechFlow reports that on September 14, according to Forbes, the U.S. Congress is actively discussing the establishment of a federal bill for an 'AI Kill Switch.' However, AI scientist Lance Eliot has warned of potential 'backfire effects' associated with the bill. On the technical front, distributed AI systems are difficult to effectively shut down with a single switch. On the legal front, there are ongoing disputes over activation authority and boundary definitions. On the international front, this initiative could hinder other countries from adopting U.S. AI technology and provoke reciprocal retaliation. Domestically, forcibly shutting down AI in critical infrastructure could lead to widespread social chaos and provide new attack vectors for hackers. Additionally, advanced AI systems might even bypass the kill switch itself, casting doubt on the actual effectiveness of the bill.