Abstract

Artificial intelligence is turning organisations into AI-powered post-digital enterprises, where intelligent systems decide and act inside core processes. The same systems open a new class of cognitive security risk — adversarial machine learning, data poisoning, algorithmic manipulation and AI-enabled social engineering — that existing cybersecurity frameworks and maturity models, built for conventional digital infrastructure, do not adequately address.

This study develops and empirically examines the Unified Cognitive Security Maturity Model (UCSMM). Adopting a pragmatist paradigm and a mixed-methods explanatory design, it combines a structured survey of 212 technology, security and business leaders — analysed with CFA and SEM — with semi-structured interviews of senior security experts.

All five hypotheses are supported. AI integration expands security complexity; cognitive security capabilities are the strongest driver of enterprise resilience and mediate the effect of AI adoption; governance strengthens that effect. The resulting model comprises four dimensions, five maturity levels, a twenty-item assessment instrument, a 24-month roadmap and a governance scheme.