The Vulnerability of Autonomous Warfare Systems

Real Clear Defense  |  Burak Oktenli

The United States military is rapidly acquiring autonomous warfare systems, including collaborative combat aircraft and uncrewed ground vehicles, without adequately measuring its forces' capability to operate when these technologies are jammed or spoofed. This rapid integration of artificial intelligence and machine-speed sensor fusion creates an exposed flank if adversaries disrupt critical satellite navigation and communication signals.

Lessons from the war in Ukraine demonstrate that electronic warfare, rather than kinetic defense, serves as the primary threat to uncrewed platforms, forcing reliance on fiber-optic controls and edge-based machine vision. While the Army's space training doctrine acknowledges these degraded environments, realistic electromagnetic denial training remains restricted to limited domestic ranges like the White Sands Missile Range. Consequently, manual reversion skills such as analog land navigation and degraded-mode gunnery are eroding unevenly across formations. Addressing this vulnerability requires establishing standardized readiness reporting for manual operations and routinely simulating network-denied conditions during home-station training exercises.

Comment

The US Army's reliance on multi-domain synchronisation during Project Convergence exercises reveals a doctrinal vulnerability where automated sensor-to-shooter links lack resilient analog redundancies. While hardened signals like M-code GPS offer localised protection against electronic attacks, they cannot fully substitute for the cognitive skills required to execute manual reversion under sustained electromagnetic denial. Consequently, the restriction of realistic jamming exercises to isolated facilities like the White Sands Missile Range prevents the institutionalisation of degraded-mode proficiency across standard combat formations.

To read in details click here