The danger of military AI isn't killer robots; it's worse human judgement (defenseone.com)

• Automation bias threatens critical human oversight.
• Cognitive load leads to over-reliance on AI decision-support.
• Command and control integrity is at risk from algorithm dependency.

"The primary threat posed by AI in military applications is not the fictional rise of 'killer robots,' but the systemic erosion of human cognitive oversight. Reliance on algorithmic outputs creates automation bias, where commanders fail to challenge flawed machine-generated insights. This degradation of human-in-the-loop accountability significantly complicates command and control, increasing the risk of unintended escalation during high-pressure combat scenarios."