UK GOVERNMENT REJECTS AI "KILL SWITCH" PROPOSAL

The UK Government has rejected proposals to create a legal "kill switch" mechanism to shut down rogue artificial intelligence systems in an emergency. The Cabinet Office stated on Friday that Britain "cannot simply turn AI off" and that blocking access to models within the UK would not prevent their development or misuse elsewhere. The rejection signals opposition to an amendment tabled to the Cyber Security and Resilience Bill by Liberal Democrat peer Lord Clement-Jones, who sought to establish statutory powers for the state to direct an emergency shutdown of autonomous models.

Lord Clement-Jones warned in House of Lords debate that security services and regulators currently lack "specific agile statutory mechanism to compel a physical or digital shutdown" if a highly capable autonomous model begins exhibiting what he termed "rogue behaviour, compound algorithmic failure or active alignment collapse". The Government stated that companies bear responsibility for developing products safely and investing in required security infrastructure, and that it would continue a "long-term, science-led approach" to emerging AI risks through its AI Security Institute. Daniel Kokotajlo, a former OpenAI researcher and co-author of the influential AI2027 research paper, stated that a single country's kill switch would be ineffective without international agreement, saying such action would "do little to protect you" against artificially intelligent systems developed elsewhere.

The Cabinet Office indicated it is exploring whether further targeted interventions may be needed to address significant risks from AI, provided any measures remain proportionate. The Government's opposition does not prevent the legislation from progressing through Parliament, though it makes passage unlikely. Prime Minister Andy Burnham has elevated AI safety minister Kanishka Narayan to the Cabinet table, signalling increased focus on the issue following warnings from safety researchers about existential risks posed by the technology.