This study addresses the persistent failure of static trust models including perimeter defenses, multi-factor authentication (MFA), and PCI-DSS compliance to detect insider threats and account takeovers (ATO) after initial authentication. Using the LANL Unified Host and Network Dataset, CERT Insider Threat Dataset v6.2, and DARPA OpTC traces, this research develops and empirically validates the Advanced Behavioral Analytics for Zero-Trust Continuous Verification (ABA-ZTCV) framework. The framework introduces a Graph–Sequence Hybrid model, Bayesian risk calibration, and bi-objective policy optimization to balance detection accuracy with user friction. Architecturally, ABA-ZTCV is structured across four adaptive layers: (1) Telemetry Fusion, integrating multi-source behavioral and identity signals; (2) Representation Learning, combining graph-based and sequential embeddings; (3) Risk and Policy Orchestration, translating calibrated anomaly scores into adaptive zero-trust controls; and (4) Feedback and Drift Management, ensuring resilience against behavioral evolution. Empirical results demonstrate that ABA-ZTCV achieves 93% detection accuracy, reduces mean time-to-detect from over 30 days to 4 days, and lowers user friction to 6%, outperforming static MFA and compliance-driven baselines. This performance was achieved through a novel Graph–Sequence Hybrid architecture coupled with bi-objective optimization that dynamically balances security sensitivity and usability. The framework operationalizes continuous verification by unifying insider threat and ATO detection in a single, adaptive loop bridging a long-standing divide between compliance requirements and real-time behavioral security. Aligning with NIST SP 800-207 and SP 800-63-4, ABA-ZTCV provides a practical and standards-compliant pathway toward resilient, continuously verified enterprise cybersecurity.
Building similarity graph...
Analyzing shared references across papers
Loading...
Emonena Patrick Obrik-Uloho
Journal of Engineering Research and Reports
Building similarity graph...
Analyzing shared references across papers
Loading...
Emonena Patrick Obrik-Uloho (Sat,) studied this question.
www.synapsesocial.com/papers/68ec51e642911f61ef8b2408 — DOI: https://doi.org/10.9734/jerr/2025/v27i101676
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: