The UK government rejects calls for an emergency “kill switch” that would allow the country to switch off dangerous AI models. The Cabinet Office, which leads on AI safety, says the UK “cannot simply turn AI off,” arguing that such a mechanism would not reliably stop harm.
The proposal, discussed in Parliament by some lawmakers in recent weeks, is framed as a legal way to restrict or halt access to AI in an emergency as fears grow about potential risks. Several outlets report the government’s concern that blocking access in the UK would not prevent the same models from being developed or misused elsewhere, limiting the effectiveness of a national shutdown.
Outlets largely align on the government’s position and rationale. While some coverage emphasizes that the decision makes passage of the legislation unlikely, other commentary highlights broader concerns that harmful systems could still be available through other jurisdictions or channels, including the possibility of continued development outside the UK. Overall, the reporting focuses on the dispute between advocates for a rapid cutoff power and the government’s view that it would not provide dependable protection.
