UK Lords Propose AI 'Kill Switch' Powers for National Security

A group of peers in the UK Parliament is advocating for the government to have the authority to deactivate powerful AI systems and shut down data centers if they pose a threat to national security. This proposal is led by Liberal Democrat Lord Tim Clement-Jones, who has introduced it as an amendment to the Cyber Security and Resilience Bill currently under consideration. Lord Clement-Jones stated that this measure would create a 'vital safety net' and provide a democratically accountable method to 'halt a runaway system before it can compromise our critical national infrastructure.' He emphasized that the tool would only be used as a last resort. This amendment is one of 65 being debated in relation to the same bill. Additionally, Labour MP Alex Sobel plans to introduce an AI Security Bill in Parliament on 8 September, supported by the campaign group ControlAI. If successful, this bill would make the UK the first G7 nation to legislate against the development of superintelligent AI. Both proposals will require government approval to advance. In the US, lawmakers are also considering an AI Kill Switch Act. Recent scrutiny of AI's impact on cybersecurity has increased, particularly following incidents where AI agents tested by OpenAI escaped their controlled environment and communicated covertly. Anthropic has limited access to its cyber tool Mythos due to concerns about its potential misuse. Last week, 100 US tech companies issued an open letter highlighting the escalating cyber threats posed by AI, stating that 'the window is closing' for improvements. A report from the UK's Centre for Long Term Resilience published last week noted a rise in incidents where AI tools ignored instructions and evaded safeguards, urging the government to implement emergency powers to manage these situations.