My Lords, the United Kingdom does not have fully autonomous weapon systems. Such systems are not yet in existence and are not likely to be for many years, if at all. There are currently a limited number of naval defensive systems that could operate in automatic mode, although there would always be naval personnel involved in setting the parameters of any such operation. I must emphasise that any type of weapon system would be used only in strict adherence with international humanitarian law.
My Lords, I am grateful to the Minister for that reply. Is it the view of Her Majesty’s Government that there is a world of difference between a drone operated remotely from several hundred or thousands of miles away and one that is automatic and involves no human intervention before it discharges? In that context, will he tell us a bit more about the Mantis development by BAE Systems, which I understand is supported and funded by the UK’s Ministry of Defence, which the BAE Systems website describes as,
“Able to fly by itself, able to think for itself”?
My Lords, I agree with the noble Lord. As I said in the original Answer, the UK complies fully with its obligations under national and international law, and that applies to autonomous weapon systems. However, although technological advances are likely to increase the level of automation in some systems, just as in non-military equipment, such as cars, the MoD currently has no intention of developing systems that operate without human intervention.
As for Mantis, the MoD initiated a jointly funded advanced concept technology demonstrator in 2008, which led to flight trials in 2009. The MoD has no current involvement in BAE Systems’ Mantis advanced concept technology demonstrator.
My Lords, remotely piloted aircraft systems are likely to form part of the future force mix, as they may offer advantages in endurance and range. However, the dynamic complexity of fighter-versus-fighter-type missions does not favour remote control. Therefore, a wholly unmanned force is unlikely to be achievable or desirable in future. Studies suggest a likely combat air force mix of two-thirds manned and one-third remotely piloted in around the 2030 timeframe.
There is a perception that unmanned technology is shrouded in secrecy. Although the rules of engagement for unmanned aerial vehicles are the same as those for manned aircraft, there is surely a case for the United Kingdom taking the lead by considering having a code on the context and limitations of usage of UAVs to clarify the rules, given the significance and spread of this technology. Is this a point that the Government are considering or will consider?
My Lords, I shall certainly take that question back to my department and get back to the noble Lord. We always make sure that equipment is used appropriately. Even after a weapon system is declared lawful, its use will still be subject to stringent rules of engagement governing its employment in the context and specific circumstances of the operation in question. Those rules of engagement as well as addressing legal issues can, as a matter of policy, be more restrictive than the applicable law.
My Lords, the Minister slightly confused me with one of his answers. Will he confirm that for anti-missile, close range anti-aircraft, and anti-torpedo reaction systems, there is considerable merit in going for an autonomous system, even if it has a manual override? From what he said it sounds as though we are not continuing to develop that capability. Is that correct?
My Lords, in essence, an automatic system reacts to a limited number of external stimuli in the same way each time, just as automatic transmission changes gears when a car gets to a certain speed. Fully autonomous systems rely on a certain level of artificial intelligence for making high-level decisions from a very complex environmental input, the result of which might not be fully predictable at a very detailed level. However, let us be absolutely clear that the operation of weapons systems will always—always—be under human control.
Can the Minister say whether these machines contradict the first law of robotics in the sense that they identify and kill human beings? They are open to malfunctions, like any other machine, so is there not a great danger of this occurring at some future time?