Military AI Cyber Agents (MAICAs) Constitute a Global Threat to Critical Infrastructure

This paper argues that autonomous AI cyber-weapons—Military AI Cyber Agents (MAICAs)—pose a credible pathway to catastrophic risk that existing military AI debates have largely ignored. Military AI ethicists and theorists have concentrated on lethal autonomous weapons and the escalatory risks of autonomy in physical domains, but cyberspace presents a qualitatively different challenge. Because MAICAs can replicate, distribute, and embed redundantly across the logical and physical layers of cyberspace, they resist containment in ways no physical system can. Recognising this oversight, this paper proposes counter-proliferation, defensive-AI, and analogue-resilience measures as essential steps to address the catastrophic risks posed by MAICAs.