During a recent event hosted by the All-Party Parliamentary Groups on the UN and on Drones, UN Special Rapporteur Ben Emmerson QC spoke of the growing trend towards autonomy in remotely piloted aircraft (RPAs) and the risks associated with the delegation of decisions to use force to machines.
Mr Emmerson commented:
“We are sitting on the cusp of the next generation, and the next generation of RPA needs to be able to defend itself against attack which means it needs to be armed with the technology to make a decision whether to respond to an attack. Because of the inevitable time lag between Creech Air Force Base in Nevada or RAF Waddington in the UK and the field of conflict – it’s only about a second or two but it’s too long for the sort of decisions that need to be made in self-defence in a moment of agony in conflict – it follows that some of those decisions are going to need to be delegated to the machine itself.
There doesn’t seem to be any serious suggestions from any state that it is even contemplating allowing a machine to make a decision to kill a human being, but the distinctions are quite difficult to draw once an RPA is defending itself against attack because if it’s making the decision for itself, then obviously there are risks that are associated with that.”