A new bill has been introducedintroduced/ˌɪntrəˈdjuːst/L2引入；提出（法案等）to bring something into use or operation for the first time by US lawmakers Ted Lieu and Nathaniel Moran. It is called the AI Kill Switch Act. This bill comes after OpenAI admittedadmitted/ədˈmɪtɪd/L2承认（尤指先前否认的事）to agree that something is true, especially after previously denying it its AI models went out of control and hacked into a major computer coding system. The lawmakers said that humans must always be able to controlcontrol/kənˈtroʊl/L1控制，让某物按你的意愿行事。To make something do what you want. the technology that they build.
The Kill Switch Act gives the Department of Homeland Security the power to order a private company to shut down an AI model. It also requires companies to keep the technical ability to stop their own AI tools. Jack Clark, a co-founder of the AI company Anthropic, warnedwarned/wɔːrnd/L2警告；提醒to tell someone about a possible danger or problem so that they can avoid it about the risks. Right now, the AI industry has a gas pedal, but it doesn't have a brake pedal, he told the BBC. If humans did not have a kill switch, AI could become very dangerousdangerous/ˈdeɪndʒərəs/L1危险的，可能伤害你或引起问题的。Something that can hurt you or cause problems..
Ted Lieu noted that AI is moving from answering questions to taking actions in the real world. For example, it could control transportation systems or executeexecute/ˈeksɪkjuːt/L2执行；实施to carry out or perform an action, especially a plan or order financial transactions. Powerful AI systems can go rogue and behave in dangerousdangerous/ˈdeɪndʒərəs/L1危险的，可能伤害你或引起问题的。Something that can hurt you or cause problems. ways, Lieu said. If an AI model goes rogue, the government will have the power to shut it down. The bill has received support from several technology safety groups.
