Options
Autonomous weapons for humanity: lessons from the Russo-Ukrainain war
Citation Link: https://doi.org/10.15480/882.17249
Publikationstyp
Book Part
Date Issued
2024-12
Sprache
English
Author(s)
TORE-DOI
Volume
19
Start Page
199
End Page
216
Citation
National Defence Academy, Austria 978-3-903548-04-6: 199-216 (2024)
Publisher
National Defence Academy, Austria
Peer Reviewed
true
ISBN of container
978-3-903548-04-6
Though autonomous weapon systems (AWS) fulfill a number of critical combat roles, and have done so for decades, ethical and legal worries have been raised in relation to recent advances in artificial intelligence (AI) and the possibility for more open-ended or versatile deployments of these systems. In particular, the potential for AWS to be used in anti-personnel roles is argued by many to present especially difficult issues ranging from concerns over the dignity of those targeted by such systems to practical worries that AWS will be incapable of making the nuanced judgments required for distinguishing combatants from civilians. However, the ongoing Russo-Ukrainian War highlights a potential response to these objections, namely that autonomous weapons may be deployed in order to further humanitarian aims by providing a means for would-be interveners to create humanitarian corridors for fleeing civilians. The argument focuses on one of the most objectionable forms of AWS, namely explicitly anti-personnel systems which aim to kill combatants, and argues that even rudimentary AWS with limited abilities of distinction could be used to bolster humanitarian aims while adding only a minimal risk to civilians. This is because object recognition in computer systems has reached a level where such systems would be able to rather reliably determine whether an individual is armed or not (and these capabilities are continually improving), and based on this criterion alone AWS could be deployed to create safe corridors of retreat for fleeing civilians. Such corridors could be marked to all sides in a conflict, their locations communicated to civilians seeking refuge, and the AWS programmed to lethally engage anyone entering the designated corridor who is armed. As the purpose of humanitarian corridors is to enable civilians to safely exit a combat zone, there is no good reason for armed units of either side to enter such a space, and so targeting based solely on whether or not one is armed could also be used as a means to show resolve in furthering humanitarian aims while taking a decidedly neutral position in the conflict itself. AWS deployed for such purposes could thus engage armed units with far less discrimination, aiming at anyone who is armed, rather than only "enemy" combatants, and the securing of such corridors through wholly unmanned systems has the added benefit that it risks no lives of would-be interveners, making participation more likely. The general simplicity of the targeting parameters for AWS deployed to such tasks (if armed, engage, otherwise, not) would furthermore ensure that the interveners could only be utilizing their autonomous weapons for purely humanitarian aims rather than as a cover for bolstering the position of one side in a conflict. The arguments further highlight a potential use-case for what is widely considered one of the most objectionable forms of AWS, presenting a counterpoint to critics' claims.
Subjects
Autonomous Weapon Systems
International Humanitarian Law
Distinction
Humanitarian Intervention
DDC Class
355: Military Science
177: Ethics of Social Relations
Publication version
publishedVersion
Loading...
Name
wood2024humanity.pdf
Type
Main Article
Size
7.76 MB
Format
Adobe PDF