Autonomous Weapon Systems (AWS)
- PDF / 222,282 Bytes
- 8 Pages / 504.567 x 720 pts Page_size
- 50 Downloads / 278 Views
Autonomous Weapon Systems (AWS) Anzhelika Solovyeva1 and Nik Hynek2 1 Institute of Political Studies, Faculty of Social Sciences, Charles University, Prague, Czech Republic 2 Department of Security Studies, Faculty of Social Sciences, Charles University, Prague, Czech Republic Keywords
Autonomous weapon systems · Artificial intelligence · Campaign to stop killer robots · International law · Revolution in warfare
Definition Autonomous weapon systems (AWS) are reusable weapon systems and smart munitions that can be differentiated from all the existing weapons by their full autonomy. It is based on (a) their ability to operate without human control or supervision in dynamic, unstructured and/or open environments; (b) their ability to engage in autonomous (lethal) decision-making, targeting and force; and (c) their ability to engage in defensive and/or offensive combat (Sharkey 2010, p. 370, 2012, p. 787; Asaro 2012, p. 690; Kastan 2013, p. 49; Open Letter 2015; Altmann and Sauer 2017, p. 118). These capabilities technologically build
upon advances in Artificial Intelligence (AI), in particular Machine Learning (ML), and especially Deep Learning (DL) and Artificial Neural Networks (ANNs) (O’Connell 2014, p. 526; Walsh 2015, p. 2; Gadiyar et al. 2019). Weapons with various degrees of autonomy are widely present on the modern battlefield; however, fully autonomous ones “do not yet exist” (Walsh 2015, p. 2). As for more advanced manifestations of autonomy, there exist remotely operated systems, including drones/unmanned aerial vehicles (UAVs) (see ▶ “Drone Warfare”) and unmanned ground and underwater vehicles. The examples include the United States (US) MQ1Predator and MQ9-Reaper UAVs, and weaponized ground robots such as the Talon SWORDS (Special Weapons Observation Reconnaissance Detection System). While they can be deployed offensively (Lele 2017, pp. 58–59) and can be lethally armed (Lucas 2014, p. 319), they are “uninhabited” rather than unmanned (Leveringhaus 2016, p. 3). This is because their primary autonomous mission is to navigate, while the selection and engagement of targets require human input (Lele 2017, p. 59). Such technologies are portrayed as operational with a human “in-the-loop” (Noone and Noone 2015, p. 28). In contrast, AWS will allow humans, besides being physically removed from the kinetic action, to become detached from decisions to fire/kill and their execution (Heyns 2013, p. 5). They will close the gap between uninhabited and unmanned warfare (Leveringhaus 2016, p. 3) and “eliminate
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2020 S. Romaniuk et al. (eds.), The Palgrave Encyclopedia of Global Security Studies, https://doi.org/10.1007/978-3-319-74336-3_636-1
2
human judgement in the initiation of lethal force” (Asaro 2012, p. 693). There have also been deployed weapon systems that are “able to identify, track and engage incoming targets on their own” (Altmann and Sauer 2017, p. 118). However, most of them are mere “extensions of electric fences” (Johnson and Axin
Data Loading...