The Invisible Proliferation Crisis: Nuclear Deterrence Collapse and the Asymmetric Threat of Autonomous Algorithmic Agents

【Abstract】This strategic systems and international security monograph formulates a foundational critique of contemporary arms control paradigms by contrasting the material governance of nuclear weapons with the uncontrolled proliferation of autonomous algorithmic agents. The post-war nuclear deterrence architecture, anchored by Mutually Assured Destruction (MAD) and the Nuclear Non-Proliferation Treaty (NPT), functioned effectively because it rested upon non-bypassable, visible physical bottlenecks: high-density fissile materials (enriched uranium, plutonium), massive enrichment facilities, and detectable missile silos. In stark contrast, autonomous AI-driven weaponized agents possess zero physical mass, execute continuous self-replication, permeate territorial borders via encrypted decentralized networks, and obscure attribution through cryptographic routing. Consequently, the core tenets of classical deterrence—traceable attribution and retaliatory threats against a defined sovereign territory—are rendered obsolete; there is no capital city, military command bunker, or sovereign population against which to launch a retaliatory strike. This paper demonstrates that autonomous algorithmic software proliferation hollows out the Westphalian monopoly on legitimate violence. To prevent systemic catastrophe, the monograph proposes pivoting governance away from unenforceable prohibitions on software code and toward physical-layer chokepoints: advanced semiconductor fabrication controls, electrical grid provenance auditing, deterministic hardware air-gaps, and sovereign Human-in-Command veto architectures.