On Explainable AI Solutions for Targeting in Cyber Military Operations

Abstract

Nowadays, it is hard to recall a domain, system, or problem that does not use, embed, or could be tackled through AI. From early stages of its development, its techniques and technologies were successfully implemented by military forces for different purposes in distinct military operations. Since cyberspace represents the last officially recognized operational battlefield, it also offers a direct virtual setting for implementing AI solutions for military operations conducted inside or through it. However, planning and conducting AI-based cyber military operations are actions still in the beginning of development. Thus, both practitioner and academic dedication isrequired since the impact of their use could have significant consequences which requires that the output of such intelligent solutions is explainable to the engineers developing them and also to their users e.g., military decision makers. Hence, this article starts by discussing the meaning of explainable AI in the context of targeting in military cyber operations, continues by analyzing the challenges of embedding AI solutions (e.g., intelligent cyber weapons) in different targeting phases, and is structuring them in corresponding taxonomies packaged in a design framework. It does that by crossing the targeting process focusing on target development, capability analysis, and target engagement. Moreover, this research argues that especially in such operations carried out in silence and at incredible speed, it is of major importance that the military forces involved are aware of the following. First, the decisions taken by theintelligent systems embedded. Second, are not only aware, but also able to interpret the results obtained from the AI solutions in a proper, effective, and efficient way. From there, this research draws possible technological and humanoriented methods that facilitate the successful implementation of XAI solutions for targeting in military cyber operations.

Modelling the Influential Factors Embedded in the Proportionality Assessment in Military Operations

“…On this behalf, Henderson & Reece, (2018) scrutinize what a 'reasonable military Commander' means in the context of proportionality assessment and consider that this assessment (i) is subjective when the person believes, (ii) is objective but unqualified when the person reasonably believes, and (iii) is objective but qualified when an expert is reasonable objective. To this end, a reasonable military Commander is a 'reasonably well-informed person in the circumstances of the actual perpetrator, making reasonable use of the information available' (Gall, 2004) which reflects the importance of military training and exercise that facilitate proper reasoning and explanations of the decisions taken (Ali, 2020;Maathuis, 2022b). Furthermore, Hayashi (2010) considers that proportionality decisions could be taken in very stressful circumstances, illustrates the importance of the military purpose that must be achieved, and points out the negative role that poor information could have for proportionality assessment.…”

The first two months in the war in Ukraine through topic modeling and sentiment analysis

“…Appraising their borderless nature and possibility of targeting both physical and cyber targets (Maathuis, 2022), the battlefield options for cyber weapons has significantly increased in recent years through altering, disturbing, damaging, influencing, or manipulating both humans and systems via the social media environment (Chen & Ferrara, 2022). Such weapons have both direct and indirect impacts on people's cognition (NATO, 2021), beliefs, and behaviors, as well as on different (inter)national security perspectives, dimensions, and aspects (Maathuis et al, 2018a).…”

Explainable AI for Healthcare 5.0: Opportunities and Challenges

“…The use of the autonomous system for defence or military operation is the same as using the autonomous vehicle and an automated system to treat any disease that may be life-threatening. Thus, EXAI is a potential tool that can capture biases in the data distribution, dependency, and inaccuracies in data balancing [40]. This provides timely information to military setups to reassess the collected data and apply data cleaning techniques to minimize the bias.…”