Britain’s government rejects a proposal to create a legal “kill switch” that would allow ministers to deactivate an AI model during an emergency. The idea was raised in Parliament as a way to respond quickly to a dangerous incident involving an AI system.
The reports say the government argues that turning AI off is not as straightforward as the proposal suggests. Officials point to practical and technical constraints, and to the challenge of defining what it would mean to shut down a specific model during an unfolding event. The “kill switch” concept is framed by supporters as a safeguard for preventing or limiting harm.
While the proposal calls for a defined mechanism and legal basis for emergency action, the government’s position emphasized limits on immediate control and the need for workable measures. The outlets do not identify a single alternative plan in detail, but the overall dispute centers on whether an emergency shutdown mechanism is feasible, enforceable, and effective under real-world conditions.
