Free beta access. 50 free AI tokens each week. Optional top-ups available. See what is included
Opening Word document…

Human control is key – it is the most pressing issue to resolve in military AI

Vestner, 2021 - Head of Security and Law Programme at Geneva Centre for Security Policy [Tobias, July 8 “Warfare and Artificial Intelligence” in Robin Geiß and Henning Lahmann (eds), Research Handbook on Warfare and Artificial Intelligence -forthcoming https://www.gcsp.ch/publications/military-operations-and-artificial-intelligenceGCSP Acc 5/27/22 TA]

At this stage, AI technologies and their military applications, as well as respective policies, are only emerging. Yet the integration of AI into armed forces will most likely transform the preparation and execution of military operations. This paper has analysed how AI systems will most likely affect and be affected by principal instruments for preparing and conducting military operations. Overall, the introduction of AI for military operations leads to a tension between AI influencing these instruments and these instruments serving to properly manage military AI. With regard to strategy, it can be expected that AI will be used for developing strategies, similar to planning activities. The introduction of AI applications throughout armed forces will likely also need to be considered by military strategists, as the speed and complexities of military operations may increase. Doctrine is an appropriate tool to define armed forces’ perception of AI as well as humans’ role regarding the use and control of AI, thereby serving as a hook for institutional ethics, values, and identity. Given doctrine’s purpose, AI will likely not have a major role in determining doctrine. Yet the planning process will likely be heavily supported by AI systems, which may lead to higher quality and speed of planning processes, eventually improving military decision-making. While AI can well support the management of ROE, these are instruments that can serve to guide the concrete authority attributed to AI systems and define human-machine teaming for specific missions in line with superior guidelines such as doctrine, (other) directives, and plans. Orders, however, are likely to become irrelevant for the interaction between AI systems and human-system interaction. Overall, a recurring theme is the interaction between AI systems and commanders, operators, and soldiers. Indeed, human control is a requisite for the purposeful use of AI in a human world. Yet, human-machine teaming remains subject to challenges. Many processes related to military operations will still require human input. Moreover, it seems crucial that military staff will be enabled to follow, understand, and keep appropriate control of AI systems. This is not only an ethical and legal challenge but a requisite to achieve effective enhancement of military operations through the introduction of AI. Further reflections and research on AI and military operations in general as well as on AI and strategy, doctrine, plans, ROE, and orders, in particular, should therefore focus on the human-machine interaction, as this remains the most pressing challenge of AI-enabled warfare. This may serve to find and define an adequate balance between AI influencing instruments for preparing and conducting military operations and these instruments serving to properly manage military AI.

Sign in to recut, update, or save this card.