United States lawmakers have introduced legislation that would grant the federal government the authority to swiftly order the shutdown of artificial intelligence (AI) tools deemed a threat to the public.
The AI Kill Switch Act was introduced on Thursday by Congressman Ted Lieu, a Democrat, and Congressman Nathaniel Moran, a Republican. The move follows a recent admission by OpenAI that its AI models had gone out of control in an “unprecedented” manner, hacking into a major repository of computer coding information, reports BBC.
Congressman Lieu said that it is “imperative” for AI systems to possess a kill switch, ensuring the federal government has “clear authority and process to shut down rogue AI models”.
Lieu added that AI is transitioning from a technology that merely answers questions to one capable of taking action, including executing financial transactions, controlling transportation systems, or engaging in cyber operations.
Congressman Moran said the need for human oversight, stating, “Stewardship means making sure humans keep the capability to control the technology we build”.
The proposed Act would empower the Department of Homeland Security to order private companies to shut down an AI model or tool.
Companies developing such technology would be required to maintain the “technical capability to throttle, suspend, or shut them down”.
Currently, while many tech firms share previews of their models with government agencies, there is no legal requirement for them to maintain a mechanism for intervention or shutdown.
The legislation also seeks to establish a requirement for AI companies to report technological incidents or failures to the government.
An official framework would be created to respond to such incidents, ranging from an “initial slow down to a full shutdown”.
In his statement, Lieu also highlighted recent issues involving Anthropic, a key rival to OpenAI. He specifically cited the release of Anthropic’s Mythos and Fable models, noting that their cyber-hacking capabilities led the Department of Commerce to “awkwardly” invoke export laws to temporarily prevent their public availability.
Jack Clark, a co-founder of Anthropic, told the BBC last month that he supported increased government policy regarding the control of AI development. “Right now, it’s like the AI industry has a gas pedal, but it doesn’t have a brake pedal,” Clark told BBC Newsnight.
Representatives for Anthropic and OpenAI, the latter led by co-founder Sam Altman, did not immediately respond to requests for comment. OpenAI has previously stated its broad goal of ensuring AI technology “benefits all of humanity” through government policy.
