SYSTEM-COGNITIVE MODELING OF INFECTION VECTORS WITH MALICIOUS CODE IN DISTRIBUTED INFORMATION ENVIRONMENTS AND FORMATION OF AN ADAPTIVE MULTILEVEL CYBER DEFENSE STRATEGY

A comprehensive analysis of the transformation of the global cyber threat landscape in the period 2024–2025 was conducted, focusing on the evolution of infection vectors through malicious code and the growing role of fileless attacks. Based on the generalization of statistical data and conclusions of leading industry reports (Verizon DBIR 2024, CrowdStrike Global Threat Report 2025), a systematic assessment of trends was conducted, indicating a fundamental shift from classic infection scenarios to interactive intrusions focused on compromising credentials and legitimate use of system components (Living-Off-the-Land). It was found that the share of fileless attacks in the structure of modern incidents exceeds 79%, which requires a revision of traditional detection and response models. Technical mechanisms for circumventing security measures were investigated, including the use of Living-Off-the-Land binaries (LOLBins), adaptive file vectors, firmware-level exploits, and physical interfaces (BadUSB). Based on the results of the analytical comparison of approaches, a multi-layered cyber defense architecture based on the principles of Zero Trust and combining content disinfection (Content Disarm and Reconstruction, CDR), behavioral monitoring (EDR/XDR), and a structured incident response protocol in accordance with the NIST SP 800-61 Rev.2 standard was proposed. Particular attention was paid to the economic feasibility of implementing multi-factor protection, where multi-factor authentication (MFA) proved to have the highest return on investment, providing over 99.9% efficiency in preventing account compromise. The results confirm the need to combine technical, organizational and behavioral protection mechanisms to form adaptive threat-resistant information environments.