Fei Su, Vladislav Chernavskikh and Dr Wilfred Wan
Introduction
Rapid developments in artificial intelligence (AI) capabilities have driven government investment in efforts to explore their applicability in military contexts, including in nuclear forces.1 The extent to which AI capabilities will be adopted by the military remains debated, but they have potential utility across nuclear weapon systems. This includes in early warning and intelligence, surveillance and reconnaissance (ISR), in nuclear command, control and communications (NC3), and in delivery systems, and also in conventional systems with counterforce potential.2 However, the use of AI in these systems would have an impact on deterrence practices and has the potential to upend strategic predictability and stability.
While reducing nuclear risks is in the collective interest of all states, there has been no discussion on establishing effective governance frameworks specifically tailored to the nexus between AI and nuclear weapons. Broader governance discussions pertaining to the use of AI in military operations have primarily focused on issues of safety, security and responsibility; these concerns are also likely to shape debates in the AI–nuclear context. However, discussions have not yet been effectively adapted to the unique challenge of nuclear forces, particularly when considering their interconnectivity with other critical technologies and domains, such as information and communications technology (ICT) and outer space.3
No comments:
Post a Comment