Cybersecurity Challenges in AI - Integrated Accounting and Auditing Systems: Empirical Evidence from Vietnam

Abstract

The rapid proliferation of artificial intelligence (AI) and digital transformation technologies in the Vietnamese business environment has profoundly altered the landscape of accounting and auditing, simultaneously introducing new and complex information security challenges. This study applies the Technology-Organization-Environment (TOE) framework to examine factors affecting Accounting Information Security (AISe) in Vietnamese enterprises during the ongoing digital transformation. The research model integrates four independent variables — Senior Management Support (TMS), Information Security Culture (SC), Quality of Accounting Information Systems (QAIS), and Cybersecurity Readiness (CSR) — and constructs the dependent variable AISe as a second-order formative construct encompassing three dimensions of the CIA triad: Confidentiality (CISe), Integrity (IISe), and Availability (AvISe). Using Partial Least Squares Structural Equation Modeling (PLS-SEM) with SmartPLS 4.0, the study analyzed data collected from 228 respondents across Vietnamese enterprises. Results show that QAIS exerts the strongest positive influence on AISe (β = 0.570, p < 0.001), followed by SC (β = 0.152, p = 0.047) and CSR (β = 0.151, p = 0.046). TMS does not directly affect AISe but exerts a strong indirect effect through SC (β = 0.777, p < 0.001). The model explains 69.4% of variance in AISe (R 2 = 0.694). These findings offer empirical evidence supporting the integrated role of technological and organizational factors in securing accounting information within Vietnam's developing digital economy. Recent developments in AI governance, including Resolution No. 57-NQ/TW (2024) and Vietnam's national AI transformation strategy (2025-2026), further underscore the urgency of these findings for policy and practice.