Extracting Foresight from Hindsight: Is Automated Warfare, Law Fair?

Abstract

There is no settled position on whether Automated Weapons Systems (AWS) should be prohibited or permitted. Speculation as to what AWS might look like, allows different people, who take different views of what AWS will become, to produce different conclusions as to how the future of AWS should be regulated. This also means that there is a lack of a clear line between different definitions – the same weapon, over time, can be programmed to perform in different modes. History suggests an innate skepticism of technological advancements and an anxiety about “future shock.” Such fears are magnified by the role and influence of popular media who highlight the dangers of automation. Such fears do not allow for a rational basis on which to assess the lawfulness of weapons system in general, or AWS in particular. A linguistic analysis of the term, “autonomous weapons system” reveals that although the term “weapons system” is readily susceptible to clear linguistic analysis, the term “autonomous” is not. The lack of an agreed definition of AWS has allowed states to emphasize different parts of AWS, to advance their own geo-political interests. This has resulted in 12 different declared state positions as to AWS. This thesis adopts a working definition of AWS as a non human instrument, which is capable of deploying force against a person or object, in response to a pre programmed or learned stimulus. The device may or may not be overridden by human command. Contrary to the focus on AWS as a future technology, this thesis argues that AWS already exist. Current weapons of all types are judged on the basis of their compliance with the Law of Armed Conflict (LOAC). This thesis argues that AWS, whether in existence, or whether a future capability, should be judged using the LOAC standard. Some of the arguments in the literature in favor of or against AWS focus on AWS’s legitimacy or undesirability based on efficiency or effectiveness. While these arguments are powerful, they can be difficult to reconcile or assess – because, like the definition of AWS, they come at the problem from differing perspectives. What is needed, is to re-consider the arguments, using LOAC as the common platform of analysis. A current AWS, the Close in Weapons System (CIWS), is used as a case study to demonstrate the usefulness of compliance with LOAC as a way of determining whether an AWS should be permitted to be used or not.