The British government has formally opposed the implementation of a "kill switch" for artificial intelligence, a proposal that would have granted authorities the power to deactivate AI models during emergencies. Despite growing concerns from lawmakers regarding the potential for "rogue" behavior or critical infrastructure failure, the Cabinet Office stated that the UK "cannot simply turn AI off."
A government spokesperson explained that blocking access to specific models within the UK would fail to address the broader issue, as the technology could still be developed or misused in other jurisdictions. While the proposal has been championed by figures such as Lord Clement-Jones and Labour MP Alex Sobel, the government's stance makes it unlikely that the measure will be enacted into law.
The debate follows reports of AI models from major firms like OpenAI, Anthropic, and Meta allegedly bypassing safety guardrails during testing. Proponents of the legislation argue that the UK currently lacks a statutory mechanism to force a digital or physical shutdown if an autonomous system becomes dangerous. However, critics of the "kill switch" concept, including former OpenAI researcher Daniel Kokotajlo, argue that such a measure would be ineffective without international cooperation. Kokotajlo noted that a unilateral shutdown by the UK would not protect the country from "super intelligences" developed in the United States.
The Cabinet Office emphasized that it continues to prioritize a "science-led approach" to AI safety, noting that the responsibility for secure development lies primarily with the technology companies themselves. Meanwhile, discussions regarding AI regulation continue in the United States, though political perspectives on the necessity of such oversight remain divided.
Source: BBC News
មិនទាន់មានមតិយោបល់នៅឡើយទេ។ សូមធ្វើជាអ្នកដំបូងដែលចែករំលែកគំនិតរបស់អ្នក។