Abstract
Legal and extra-legal factors related to the development of international humanitarian law (IHL) support the exercise of strategic restraint during armed conflicts. We claim, however, that technological developments in the area of military artificial intelligence (AI) fundamentally do away with many of the technological constraints that limited the capacity of militaries to fully utilize the “license to kill” afforded to them by IHL. Such a violence-enhancing effect can become even more dramatic in cases where other restraining factors are less effective. The upshot of our analysis is that the implicit assumption that IHL would be applied in a context of strategic restraint might no longer hold true. Part II of the article discusses the role of strategic restraint in IHL. In Part III, we consider three principal extra-legal sources of strategic restraint—moral considerations, practical considerations, and technological constraints—and explain why the use of algorithmic decision support systems (DSS) has the potential to seriously erode their restraining effect during armed conflicts. In Part IV, we illustrate the inter-relations between the different sources of strategic restraint by way of referring to the recent conflicts in Ukraine and Gaza. In Part V, we consider possible responses to what we consider to be a paradigmatic shift in the application of IHL—introduction of limits on the use of military AI—including baking into the technology elements of restraint—and changes in the contents of IHL and jus ad bellum that could compensate for the erosion of strategic restraint. Part VI concludes.
html
