The UK Ministry of Defence is reviewing a policy change to let AI-powered weapons select and strike targets independently in limited high-stakes scenarios, like overwhelming data or sudden threats. This would mark a departure from the current 'human in the loop' rule, driven by advances in drone warfare seen in Ukraine and NATO's push for quicker decisions. While supporters say it could save lives by processing data humans can't, critics warn of risks like misidentified targets and rising civilian casualties, echoing stalled UN talks on autonomous weapons.
