59 2 EXTENSIVE LITERATURE REVIEW combines systems thinking and systems engineering as conceptual tools to frame the commonalities between these two LoAs. A third approach to operationalise Meaningful Human Control is presented by Cavalcante Siebert et al. (2023) who are proposing four actional properties for AI-based systems under Meaningful Human Control to bridge the gap between philosophical theory and engineering practice. In their reflection on their work the authors highlight that ‘Meaningful human control is necessary but not sufficient for ethical AI.’ (Cavalcante Siebert et al., 2023, p. 252). The authors amplify this by stating that for a human-AI system to align with societal values and norms, Meaningful Human Control must entail a larger set design objectives which can be achieved by transdisciplinary practices. However, the wider conception of the control loop mentioned above does not incorporate the social institutional and design dimension at a governance level. The governance level is the most important level for oversight and needs to be added to the control loop, because accountability requires strong mechanisms in order to oversee, discuss and verify the behaviour of the system to check if its behaviour is aligned with human values and norms. Institutions and oversight mechanisms need to be consciously designed to create a proactive feedback loop that allows actors to account for, learn and reflect on their actions. Therefore, we look at an oversight mechanism to connect the technical, socio-technical and governance perspective of control which may ensure solid controllability and accountability for the behaviour of Autonomous Weapon Systems. 2.12 HUMAN OVERSIGHT Several scholars mention that an oversight mechanism is needed in order to hold an actor accountable (Caparini, 2004; Schedler, 1999; Scott, 2000). West and Cooper (1989: in (Pelizzo, Stapenhurst, & Olson, 2006)) mention two reasons for oversight in the political system: (1) it can improve the quality of policies or programs and (2) when policies are ratified by the legislative branch, they obtain more legitimacy. The oversight mechanism can be implemented as an ex post review process or a mechanism for either ex post of ex ante supervision (Pelizzo et al., 2006). According to Goodin (1995) responsibility needs supervisory action in that A has to see to it that X is achieved. He states that ‘… require[s] certain activities of a self-supervisory nature from A. The standard form of responsibility is that A see to it that X. It is not enough that X occurs. A must also have “seen to it” that X occurs. “Seeing to it that X” requires, minimally, that A satisfy himself that there is some process (mechanism or activity) at work whereby X will be brought about; that A check from time to time to make sure that that process is still at work, and is performing as expected; and that A
RkJQdWJsaXNoZXIy MjY0ODMw