The Guardian reports that the UK will soon issue “a new doctrine designed to calm concerns about the development of killer robots.” While it is beyond doubt that any such “new doctrine” will be designed to calm our concerns, based on the Guardian report it is unclear that anything new has been promised.
The US and UK have been playing the same basic game: posit fully autonomous weapons as a Kantian ideal of human-like intelligence and wanton killing for reasons of their own — something no one ever wanted — and disavow any intention of ever creating such monsters, far beyond current technology anyway. Meanwhile, plow ahead on every front of the robot arms race under the assurance that they are not creating the mythical out-of-control killer robots.
According to the Guardian, the MOD has pledged that “UK policy is that the operation of weapons will always be under control as an absolute guarantee of human oversight, authority and accountability.The UK does not possess fully autonomous weapon systems and has no intention of developing them.”
While placing this in quotes, The Guardian does not explicitly say who it is quoting. The closest match seems to be a letter to Richard Moyes of Article 36 from Paul Larkin of the Counter Proliferation and Arms Control Centre.
That letter reiterates the view that “The UK defines such systems as those capable of understanding higher level intent and direction allied to a sophisticated perception of its environment…” in other words, defining AWS as systems with approximately human-level intelligence, systems everyone will agree don’t exist yet, if ever.
The letter also reiterates that, even with AWS defined as far-off, human-level AI-driven killer robots, “The UK does not support a pre-emptive ban on such systems.” It argues that a ban would “stifle research” into the “legitimate non-lethal advantages to increasingly autonomous technology in the future.” It views “existing International Humanitarian Law [as] sufficient to control and regulate Lethal Autonomous Weapons Systems.”
In any event, if we take the Guardian’s unattributed quote as a statement of policy, there is no indication that it deviates from previous British policy, which like the essentially equivalent American policy, draws attention to mythical killer robots that nobody wants and are beyond current technology, while casting a veil over everything the military is actually doing or might potentially want to do, calling it merely “automatic” or “semi-autonomous” and not of concern.