Skal vi forbyde dræberrobotter?

Dødelige autonome våbensystemer kræver omhyggelig overvejelse, men fremtidens mareridtsscenarier vil ikke blive til virkelighed når som helst snart, siger en militæretiker fra UNSW Canberra.

Udtrykket 'dræberrobotter' fremmaner billeder af sci-fi-scenarier, hvor krige udkæmpes af Terminator-lignende soldater, men ifølge UNSW Canberra militære etiker Deane-Peter Baker, er det ikke helt så skræmmende eller filmisk.

Faktisk kan dræberrobotter eller dødelige autonome våbensystemer (LAWS) faktisk redde liv på slagmarken.

Lektor Bakers seneste bog, "Skal vi forbyde dræberrobotter?", bygger på hans erfaringer fra International Panel on the Regulation of Autonomous Weapons (IPRAW) fra 2017-19.

IPRAW er et internationalt netværk af forskere, der har til opgave at yde upartisk vejledning til de nationale delegationer, der er engageret i FN-debatten om, hvorvidt der skal forbydes eller reguleres LOVE.

"Denne bog er mit forsøg på at samle mine synspunkter om dette emne, som er opstået fra min tid som IPRAW paneldeltager og andet efterfølgende politik-fokuseret arbejde om dette emne," A/Prof. sagde Baker.

Han forklarede, at der er to hovedargumenter for at forbyde LOVE. Man fokuserer på de potentielle konsekvenser af at tillade LOVE at blive brugt i krig.

"For eksempel er modstandere bekymrede for, at LAWS ikke vil være i stand til at fungere inden for grænserne af loven om væbnet konflikt," A/Prof. sagde Baker.

"Bekymringen her er, at de vil bruge magt på en vilkårlig eller uforholdsmæssig måde.

"Den anden hovedtype argument er, at bortset fra konsekvenserne er det simpelthen grundlæggende forkert at tillade en maskine at træffe valget om at dræbe et menneske."

Ifølge A/Prof. Baker, mindre udviklede stater har en tendens til at gå ind for et forbud, mens magtfulde og teknologisk avancerede stater ikke er særlig støttende.

"Fortalere for LOVE hævder, at disse systemer kan redde liv på en række måder," sagde han.

"For example, there is the claim that robots can be sent to do 'dull, dangerous and dirty' jobs without having to risk a human soldier, sailor or aviator—far better for a machine to get destroyed than for a member of the armed forces to be killed or maimed.

"The other main type of argument is that, consequences aside, it's simply fundamentally wrong to allow a machine to make the choice to kill a human being."

"They also argue that LAWS will be less prone to using indiscriminate force, because they don't get scared, angry or confused in the way that human combatants can in the midst of combat."

A/Prof. Baker said there is also the argument that an international ban will not prevent malign actors from developing and using these systems, so we should not hand them a significant operational advantage by adopting a ban and disallowing ourselves from using them.

So, will we find ourselves in that Terminator situation any time soon?

"We're a long way from that happening, if indeed it ever does!" A/Prof. Baker assured.

"I think there's no doubt that we will start to see more and more lethal autonomous weapons participating in wars—the UN believes we have already seen the first humans to be killed by autonomous weapons, in the Yemen conflict. But it's my view that they will be unlikely to play much more than a supplementary role for some time to come."

In the medium term, he said highly sophisticated systems will be very expensive and therefore rare, while simple autonomous systems will be constrained by limited capability.

"Over the longer term we will start to see more sophisticated systems becoming more affordable and therefore more prolific, and the simpler systems will themselves become more capable," A/Prof. Baker said.

He hopes that readers of the book will come away with a clearer understanding of the arguments that have been raised in favor of a ban on killer robots.

"Even if they don't agree with my conclusion, hopefully their thinking will have been challenged and their views sharpened in the process."