Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Killer Robots - Autonomous Weapons and Their Compliance with IHL

Andersson, Cecilia LU (2014) JURM02 20141
Department of Law
Abstract
The pursuit of weapons which distance the soldier from the actual battlefield has been going on ever since the transition from the waging of war using short blades, to the waging of war using bow and arrow. Today, that ambition has reached an almost completion with the ever-increasing number of unmanned, remote-controlled vehicles that are rapidly becoming the most common and prominent method of waging wars. Political incentives of cutting costs of warfare and sparing the lives of soldiers create the last push towards full autonomy. The emergence of increasingly autonomous weapons (AWs) has already generated a heated debate on the legality of these weapons, and two very polarized sides can be easily discerned.

The purpose of this... (More)
The pursuit of weapons which distance the soldier from the actual battlefield has been going on ever since the transition from the waging of war using short blades, to the waging of war using bow and arrow. Today, that ambition has reached an almost completion with the ever-increasing number of unmanned, remote-controlled vehicles that are rapidly becoming the most common and prominent method of waging wars. Political incentives of cutting costs of warfare and sparing the lives of soldiers create the last push towards full autonomy. The emergence of increasingly autonomous weapons (AWs) has already generated a heated debate on the legality of these weapons, and two very polarized sides can be easily discerned.

The purpose of this thesis is to examine and analyze this debate, to look into the arguments put forth regarding the legality or illegality of autonomous weapons, and examine where the positions are in the debate. Focus is on the three fundamental principles in International Humanitarian Law (IHL): distinction, proportionality and precaution, and I discuss the arguments in both directions. Proponents often claim the ability of AWs to comply with IHL, with the development of sensors, algorithms, software and artificial intelligence (AI), which would allow the machine to satisfactorily distinguish between civilians and combatants, carry out proportionality assessments and to take the required precautions in its actions. Opponents instead argue that the development of AI has overpromised before, that sensors could never be able to distinguish between civilians and combatants in a contemporary battlefield and that proportionality and precaution assessments require a contextual understanding that only humans are capable of. The fundamental disagreement seems to lie in the uncertainty of the development of the software and technology, and the capability of machines to perform as well, or better than, humans. The issue of accountability is also examined in terms of what happens with the responsibility for breaches of IHL when we have assigned the task of targeting and firing, essentially, the life-and-death decision, to a machine. Different propositions such as placing the accountability onto the commander, programmer, manufacturer or even the machine itself are discussed. Issues relating to the moral and ethical aspects of changing the agents of war from humans to robots are also examined, and the possible consequences this might entail – both from a separate moral perspective and as part of the legality assessment, in terms of what would happen with the applicability of IHL if we would change the agents in war.

After having examined the debate on legality of AWs, some concluding remarks are drawn on what we are to do with the debate in the near future, where I present some of the more prominently discussed ways forward in terms of handling the emergence of these weapons. Finally, I end with some of my own reflections on what I have found in my analysis of the current debate, and what I believe are the more important aspects to continue discussing in the ongoing debate on the legality of autonomous weapons. (Less)
Abstract (Swedish)
Jakten på vapen som distanserar soldaten från själva slagfältet har pågått ända sedan övergången från krigsföring med knivar till krigsföring med pil och båge. Idag har denna ambition närapå nått fullständighet med det ständigt växande antal obemannade, fjärrstyrda farkoster som snabbt håller på att bli den vanligaste och mest framstående metoden att föra krig. Politiska incitament såsom att kapa kostnader av krig och att spara soldaters liv innebär den sista knuffen mot full autonomi. Framväxten av alltmer autonoma vapensystem har redan genererat en passionerad debatt om lagligheten av dessa vapen, och två väldigt polariserade sidor är enkelt urskiljbara.

Syftet med det här arbetet är att undersöka och analysera den här debatten, att... (More)
Jakten på vapen som distanserar soldaten från själva slagfältet har pågått ända sedan övergången från krigsföring med knivar till krigsföring med pil och båge. Idag har denna ambition närapå nått fullständighet med det ständigt växande antal obemannade, fjärrstyrda farkoster som snabbt håller på att bli den vanligaste och mest framstående metoden att föra krig. Politiska incitament såsom att kapa kostnader av krig och att spara soldaters liv innebär den sista knuffen mot full autonomi. Framväxten av alltmer autonoma vapensystem har redan genererat en passionerad debatt om lagligheten av dessa vapen, och två väldigt polariserade sidor är enkelt urskiljbara.

Syftet med det här arbetet är att undersöka och analysera den här debatten, att titta på de argument som förs fram gällande lagligheten eller olagligheten av autonoma vapen, och att undersöka var positionerna står i debatten. Fokus ligger på de tre grundläggande principerna i internationell humanitärrätt (IHL): distinktion, proportionalitet och försiktighet, och jag diskuterar argumenten i båda riktningarna. Förespråkarna framhäver ofta förmågan hos autonoma vapen att efterleva reglerna i IHL, genom utvecklingen av sensorer, algoritmer, mjukvara och artificiell intelligens (AI), vilket skulle göra det möjligt för maskinen att på ett tillfredsställande sätt skilja mellan civila och kombattanter, genomföra proportionalitets-bedömningar samt att företa nödvändiga försiktighetsåtgärder i sina aktiviteter. Motståndarna menar istället att utvecklingen av AI har lovat för mycket förut, att sensorer aldrig skulle kunna skilja mellan civila och kombattanter i ett nutida krigsfält och att bedömningar av proportionalitet och försiktighetsåtgärder kräver en kontextuell förståelse som endast människor kan klara av. Den grundläggande meningsskiljaktigheten verkar ligga i ovetskapen om utvecklingen av mjukvara och teknologi, och förmågan hos maskinerna att utföra uppgifter lika bra som, eller bättre än, människor. Frågan om ansvar undersöks också gällande vad som händer med ansvaret för överträdelser av IHL när vi överlåter uppgiften av att sikta och avfyra, i allt väsentligt, liv och död-beslut, till en maskin. Olika förslag om var ansvaret ska placeras, såsom på befälhavaren, programmeraren, tillverkaren eller till och med på maskinen själv, diskuteras. Frågor som relaterar till de moraliska och etiska aspekterna av att byta ut agenterna i krig från människor till robotar undersöks också, och de möjliga konsekvenser detta innebär – både från ett separat moraliskt perspektiv, men också som del av laglighetsbedömningen, beträffande vad som händer med tillämpligheten av IHL om vi byter agenterna i krig.

Efter att ha undersökt debatten om laglighet av autonoma vapen drar jag några slutsatser om hur vi ska fortsätta debatten i den nära förestående framtiden, där jag presenterar några av de mest diskuterade möjliga vägarna framåt när det gäller att hantera framväxten av dessa vapen. Slutligen avslutar jag med några egna reflektioner om vad jag har kommit fram till i min analys av debatten, och vad jag tror är de viktigaste aspekterna att bära med sig i den fortsatta debatten om lagligheten av autonoma vapen. (Less)
Please use this url to cite or link to this publication:
author
Andersson, Cecilia LU
supervisor
organization
course
JURM02 20141
year
type
H3 - Professional qualifications (4 Years - )
subject
keywords
autonoma vapen, international humanitärrätt, robots. Folkrätt, autonomous weapons, Public international law, International humanitarian law, robotar.
language
English
id
4449496
date added to LUP
2014-06-12 09:07:09
date last changed
2014-06-12 09:07:09
@misc{4449496,
  abstract     = {{The pursuit of weapons which distance the soldier from the actual battlefield has been going on ever since the transition from the waging of war using short blades, to the waging of war using bow and arrow. Today, that ambition has reached an almost completion with the ever-increasing number of unmanned, remote-controlled vehicles that are rapidly becoming the most common and prominent method of waging wars. Political incentives of cutting costs of warfare and sparing the lives of soldiers create the last push towards full autonomy. The emergence of increasingly autonomous weapons (AWs) has already generated a heated debate on the legality of these weapons, and two very polarized sides can be easily discerned. 

The purpose of this thesis is to examine and analyze this debate, to look into the arguments put forth regarding the legality or illegality of autonomous weapons, and examine where the positions are in the debate. Focus is on the three fundamental principles in International Humanitarian Law (IHL): distinction, proportionality and precaution, and I discuss the arguments in both directions. Proponents often claim the ability of AWs to comply with IHL, with the development of sensors, algorithms, software and artificial intelligence (AI), which would allow the machine to satisfactorily distinguish between civilians and combatants, carry out proportionality assessments and to take the required precautions in its actions. Opponents instead argue that the development of AI has overpromised before, that sensors could never be able to distinguish between civilians and combatants in a contemporary battlefield and that proportionality and precaution assessments require a contextual understanding that only humans are capable of. The fundamental disagreement seems to lie in the uncertainty of the development of the software and technology, and the capability of machines to perform as well, or better than, humans. The issue of accountability is also examined in terms of what happens with the responsibility for breaches of IHL when we have assigned the task of targeting and firing, essentially, the life-and-death decision, to a machine. Different propositions such as placing the accountability onto the commander, programmer, manufacturer or even the machine itself are discussed. Issues relating to the moral and ethical aspects of changing the agents of war from humans to robots are also examined, and the possible consequences this might entail – both from a separate moral perspective and as part of the legality assessment, in terms of what would happen with the applicability of IHL if we would change the agents in war.

After having examined the debate on legality of AWs, some concluding remarks are drawn on what we are to do with the debate in the near future, where I present some of the more prominently discussed ways forward in terms of handling the emergence of these weapons. Finally, I end with some of my own reflections on what I have found in my analysis of the current debate, and what I believe are the more important aspects to continue discussing in the ongoing debate on the legality of autonomous weapons.}},
  author       = {{Andersson, Cecilia}},
  language     = {{eng}},
  note         = {{Student Paper}},
  title        = {{Killer Robots - Autonomous Weapons and Their Compliance with IHL}},
  year         = {{2014}},
}