ANALYSIS OF THE CONSEQUENCES OF ARTIFICIAL INTELLIGENCE INTEGRATION FOR CHEMICAL, BIOLOGICAL, RADIOLOGICAL AND NUCLEAR (CBRN) RISK DETERRENCE

Abstract

The article examines the consequences of integrating artificial intelligence technologies into the system of deterrence and countering chemical, biological, radiological and nuclear (CBRN) threats amid a qualitative transformation of the global security environment. The conceptual framework for applying artificial intelligence, machine learning and autonomous decision-making systems in the security context is analysed, with a taxonomy of AI applications in the CBRN domain delineated: predictive analytics, computer vision, natural language processing for open-source intelligence and autonomous robotics for decontamination. The CBRN threat landscape of the 2020s is explored with emphasis on the Russo-Ukrainian war, including chemical incidents along the front line (confirmed by the OPCW report on CS agent in Dnipropetrovsk region), radiation risks around the Zaporizhzhia Nuclear Power Plant (destruction of monitoring stations within the thirty-kilometre zone) and biosecurity challenges amid destroyed infrastructure. Empirical data on the effectiveness of AI detection systems for chemical agents with accuracy of 96–99 percent, radionuclide identification above 95 percent and biosurveillance based on open-source intelligence are systematised, including the EPIWATCH system experience in Ukraine. It is substantiated that the effectiveness of AI detection is conditional and depends on data quality, interoperable architecture and personnel readiness, while laboratory performance inevitably declines under combat conditions. The ethical and legal dimension of integration is studied through the prism of the meaningful human control concept, NATO six principles of responsible AI use and dual-use risks of generative models, notably the MegaSyn case of VX nerve agent generation. The Ukrainian context is analysed, including the SESU regulatory framework, experience of EPIWATCH and SaveEcoBot systems, and prospects for adaptation to the EU CBRN Action Plan and NATO standards. A phased strategy for AI integration into the national CBRN deterrence system is formulated taking resource constraints and armed conflict conditions into account. The author argues for aggressive AI deployment in analytical and detection functions while preserving human oversight over all critical decisions in the CBRN domain.