Artificial intelligence and machine learning techniques are being developed to improve decision making around the resort to force. These technologies are valued for their capacity to rapidly collect ...and analyse big data, model unique courses of action, offer probabilistic recommendations and predictions regarding the type and degree of force required, and evaluate the benefits, risks, and costs of action and inaction. Those concerned with these developments highlight the possibility of automation bias in human-machine teaming, and the potential de-skilling of individual and institutional decision making. This concern is valid, but too narrow in scope. In addition to human knowledge and skill, wisdom is imperilled by the growing technification of violence and war. Drawing on the lessons of tragedy, I argue that the speed, inflexibility, and false confidence of algorithmically assisted decision making cultivates an insensitivity to the tragic qualities of violence. This dulling of the tragic imagination is likely to lead to more imprudent and immoral uses of force, not less.
Artificial Intelligence (AI) and machine learning (ML) are rapidly evolving and have already had major impacts on military capabilities in the battlefield, making new kinds of tools and tactics ...available. A less examined area of application for AI in a military context, however, is its impact on human strategic decision making. This article focuses on the more subtle cognitive influences of AI and how they can be strategically deployed to aid decision making around the state-level resort to force, in particular. I will argue that AI-driven technologies can be used to improve certain critical cognitive resources (e.g. memory, planning, mind-modelling, etc.) of decision makers, thereby providing valuable strategic advantages to those actors who use them successfully. At the same time, I will also caution against the risks of human decision makers becoming overly reliant on AI-support systems. Both the potential advantages and risks are areas that demand further study and consideration.
Abstract
This article examines how a cyber-operation that has consequences similar to a kinetic or physical attack — causing serious loss of life or physical damage — could be encompassed within the ...crimes prosecuted before the International Criminal Court (ICC). It explains when and how such a cyber-operation could fall within the ambit of the ICC’s crimes — genocide, crimes against humanity, war crimes, and the crime of aggression. The article additionally acknowledges some limitations as to which cyber-operations would be encompassed, given the ICC’s gravity threshold as well as the potential difficulty of attributing conduct to a particular suspect through admissible evidence that could meet the requirement of proof beyond a reasonable doubt. Notwithstanding such limitations, increased awareness of the previously largely overlooked potential of the Rome Statute to cover certain cyber-operations could potentially contribute to deterring such crimes.
Cyber operations are becoming an increasing part of armed conflict. This article assesses whether cyber operations against data during an armed conflict could amount to a war crime in the Statute of ...the International Criminal Court. It unpacks the plausibility of computer data being included in the categories of 'object' and 'property' in the Statute, showing there is no doctrinal and jurisprudential unanimity in either case. However, the Court can and should take a wide view of when tangible objects are affected in a legally relevant way by attacks on or through data. Considering this question forces us to reflect about the proper interpretation of the Statute in light of the principle of legality, and about whether and how the Statute will be able to 'keep up' with new forms of warfare.
Up to now, the Chinese government has only made very general comments on the application of international humanitarian law to cyberspace. There are indeed Chinese academic papers concerning this ...issue, but the discussion of the principle of distinction is limited both in length and in academic depth. Compared with the West, research by Chinese scholars on this topic is still in a relatively preliminary stage. At present, there is no specific deconstruction or clarification of the application of the principle of distinction in cyberspace in Chinese academia. As the first paper written by Chinese scholars specifically devoted to this question, this piece provides a different perspective by injecting the positions of Chinese officials and the views of Chinese scholars. The authors aim to clarify whether the existing rules are still completely applicable in the cyber context, and if needed, to find out what kind of improvements and clarifications can be made. Weighing in on these debates, we argue that despite the potential technical challenges and uncertainties, the principle of distinction should be applied to cyberspace. It should also be carefully re-examined and clarified from the standpoint of preventing over-militarization and maximizing the protection of the interests of civilians. For human targets, the elements of combatant status identified in customary international law and relevant treaties are not well suited to the digital battlefield. Nevertheless, cyber combatants are still obligated to distinguish themselves from civilians. In applying the principle of distinction, we argue that it makes more sense to focus on substantive elements over formal elements such as carrying arms openly or having a fixed distinctive sign recognizable at a distance. In interpreting “direct participation in hostilities”, the threshold of harm requires an objective likelihood instead of mere subjective intention; the belligerent nexus should be confirmed, and the causal link should be proximate. Applying the “cyber kill chain” model by analogy helps us to grasp the whole process of direct participation in hostilities during cyber warfare. For non-human targets, all military objectives must cumulatively fulfil both the “effective contribution” and “definite military advantage” criteria, which are equally indispensable. The same requirements apply to dual-use objects. Furthermore, certain data should fall within the ambit of civilian objects.
