Member

The core argument advanced by ControlAI and leading safety researchers is simple yet profound: you cannot govern what you cannot legally stop. At present, the UK finds itself in a state of dangerous legal exposure. I agree with the noble Lord, Lord Holmes, that this is not all about frontier models but, if a highly capable autonomous frontier model, whether hosted in a UK data centre or integrated across our critical infrastructure, begins exhibiting rogue behaviour, compound algorithmic failure or active alignment collapse, our security services and regulators possess no specific agile statutory mechanism to compel a physical or digital shutdown. Relying on slow-moving administrative notices or voluntary developer good will in a minutes-critical national emergency is not a sufficient strategy.