CYIL vol. 9 (2018)

LUDMILA HALAJOVÁ

CYIL 9 ȍ2018Ȏ

Bhuta, Nehal (eds.) Autonomous Weapons Systems: Law, Ethics, Policy Cambridge: Cambridge University Press, 2016, 422 p.

[Autonomní zbraňové systémy: Právo, Etika, Politika] The book edited by Nehal Bhuta, Susanne Beck, Robin Geiß, Hin-Yan Liu and Claus Kreß is the first comprehensive collection of texts on the topic of Autonomous Weapons Systems (“AWS”), i.e. “ weapons systems that , once activated, can select and engage targets without further intervention by a human operator ”. 1 Although AWS in the sense of this definition do not exist yet and are not likely to be developed and deployed on the battlefield in the near future, the topic has gained a lot of attention in the recent years and it is increasingly debated on the academic ground as well as among various stakeholders. 2 The book contains contributions by twenty-one experts from the field of international law, robotics, politics, sociology and philosophy. The editors divided the collection into seven parts. The introductory remarks were written by Christof Heyns , a former UN Special Rapporteur on extrajudicial, summary or arbitrary executions, who is one of the most vocal proponents of the ban on the development and deployment of AWS. Heyns draws on his previous works on the topic and briefly sums up the state of the debate on AWS. The following five parts represent the substantive core of the book. The contributions focus on three thematic clusters of issues related to AWS, which have been identified by the authors as the most problematic aspects of this emerging technology. These are the notion of machine autonomy, the concept of human dignity, and the accountability for the actions of AWS. According to the editors, these aspects have been often neglected in the past academic works on the topic. The second part of the book comprises three contributions from experts in robotics, informatics, computer engineering and sociology, which focus on the notion of autonomy in weapon systems. Noel Sharkey sets out his concept of five levels of supervisory control that humans can exercise over computerized weapons systems. He also makes a distinction between the automatic and deliberative reasoning, describing their strengths and weaknesses. The next chapter, written by Giovanni Sartor and Andrea Omicini , analyses machine autonomy from the perspective of informatics. The authors describe three dimensions of autonomy that need to be present in a system in order for it to be autonomous. Those are independence (operating without external intervention), cognitive skills (performing complex actions such as information acquisition, analysis and decision adoption and implementation), and cognitive architecture (possession of direction and purpose and the capacity to adapt to changing circumstances and to achieve goals by selecting the appropriate means based on the inputs from the environment). Lucy Suchman and Jutta Weber close this part of the book 1 The most commonly used definition of AWS. See US Department of Defense, Directive 3000.09, ‘Autonomy in weapon systems’ (21 November 2012) accessed 8 May 2018, 13. 2 See for example the discussion at the first and second meeting of the Group of Governmental Experts on lethal autonomous weapons systems established by the High Contracting Parties to the Convention on Certain Conventional Weapons or the previous informal meetings of experts on the topic at .

422