Escalation by Algorithm, Restraint by Architecture: Pakistan's Military AI Divergence

Abstract

Debates on military artificial intelligence (AI) remain skewed toward great power dynamics, oscillating between techno-optimist celebrations of speed and techno-pessimist warnings of collapse. This binary overlooks how middle powers, operating under nuclearized rivalry and asymmetric sanction risk, embed restraint into organizational and technical practice. This article develops Systems Restrained Realism (SRR), a framework that extends defensive realism into the machine age by theorizing restraint as a deliberate doctrinal posture rather than a symptom of incapacity. Using Pakistan as a critical case, the study draws on expert interviews, procurement manuals, and UN submissions to demonstrate how restraint is operationalized through latency as doctrine, embedded organizational oversight, and localized training regimes designed to mitigate classifier fragility. The findings reveal that while India's accelerationist AI trajectory projects capability and ambiguity, Pakistan engineers restraint into systems and decision loops, externalizing it through normative signaling at UN forums. This posture highlights a structural asymmetry: Great powers can afford AI misfires under the banner of innovation, while middle powers face punitive scrutiny for errors, incentivizing opacity over transparency. By foregrounding SRR, the study challenges dominant narratives that equate restraint with weakness and automation with stability. It argues that in an era of machine-speed conflict, survival may hinge not on what states automate but on what they refuse to.