Un exposé par Jean-Baptiste Jeangène Vilmer, Boursier Postdoctoral Banting.
(en anglais seulement) The US Congress anticipated that “by 2015, one-third of the operational ground combat vehicles will be unmanned”, a change described as a "revolution in the art of land warfare". At present, these armed robots are remotely controlled by humans, and the ethical issues involved are already complex. On one hand, robots have certain advantages: the ability to save lives (those of the men they replace) and a superior performance at a lower cost if we take the example of UCAVs. On the other hand, they raise serious concerns including a reduced capacity for discernment and hence, discrimination (jus in bello); a tendency towards the sanitarization of war, giving the impression that it is “cleaner” than it really is; and, as a result, a lowering of the threshold for the purposes of resorting to armed violence (including possibly at the level of the jus ad bellum). The situation is further complicated with the development of autonomous models with a capacity for “judgment” and decision-making, for example to shoot at targets without human intervention.
The interesting question here is how current development programs will incorporate the principle of discrimination between combatant and noncombatant, and the principle of proportionality, all of which are not objective measures but results of human judgment. In the case of a mistaken shot fired by an autonomous robot, who should bear responsibility? While researchers are thinking of ways to treat robots as legal quasi-agents, they are not yet, and probably will not be subjects of international humanitarian law for many years to come.