The Algorithmic Brink: Ethical Governance and Escalation Risk in AI- Enabled Nuclear Systems across Space and Conflict Domains

Abstract 
The growing use of artificial intelligence (AI) 
in nuclear command, control, and 
communications (NC3), especially in space
based security systems, marks a significant 
shift in how strategic decisions may be made. 
AI systems can improve information 
processing and operational speed, but they also 
create new risks that may undermine strategic 
stability. This study explores the ethical, 
technical, and strategic challenges that arise 
when nuclear decision-making relies on 
autonomous or semi-autonomous systems. 
Drawing on deterrence theory, research on AI 
safety, studies of human–machine interaction, 
and existing space law, the paper identifies 
three primary sources of risk: shortened 
decision timelines, ambiguous responsibility 
for decisions, and weaknesses in sensor 
reliability. To demonstrate how these factors 
interact, the study presents a stochastic 
escalation model showing how fast-paced 
machine interactions can raise the risk of 
unintended conflict in uncertain conditions. 
The paper concludes by outlining a Human
Centric Heuristic (HCH) governance model 
that emphasizes sustained human control while 
still supporting timely operational decisions.