This master thesis addresses the vulnerabilities of the use of artificial intelligence in drones used for military-defence purposes. Even though AI technologies are nowadays increasingly gaining in their utility and thus value and are playing a growing role in modern defence systems, understanding the limitations, risks and vulnerabilities of new AI technologies is crucial for their further use or further integration in military-defence systems. This master thesis explores the vulnerabilities of AI through three key vulnerability factors: developmental-technological, human and formal-legal. Within the technological perspective, vulnerabilities related to the robustness of algorithms, resilience to cyber-attacks and the complexity of AI systems are analysed. The human factor addresses issues of trust, operational familiarity, control and accountability in the use of AI in military operations. The formal-legal aspect focuses on legal uncertainties, the dilemmas of international humanitarian law and the ethical issues related to autonomous decision-making in warfare. The above is supported by an analysis of actual practical cases, a review of relevant international law, and the findings of these are further supported by an expert interview. The master’s thesis contributes to a broader understanding of the challenges posed by digitalisation and the military space and provides a foundation for further research and the design of regulatory frameworks.
|
