Adoption Dynamics and Implementation Strategies of Machine Learning for Cybersecurity Threat Detection in the E-Commerce Sector

Abstract

The rapid growth of e-commerce has intensified cybersecurity challenges, as organizations face increasingly complex and adaptive threats that traditional detection systems struggle to manage. Although machine learning (ML) offers predictive and adaptive capabilities for intrusion detection, many firms experience a persistent gap between adoption intention and practical implementation. This study addresses that gap by investigating the organizational, technological, and governance factors that influence how e-commerce enterprises adopt and operationalize ML-based cybersecurity systems. It integrates theoretical perspectives from the Unified Theory of Acceptance and Use of Technology (UTAUT), Technology Organization Environment (TOE), and Diffusion of Innovations (DOI) frameworks to explain how capacity and compliance conditions translate adoption intent into sustainable deployment. Using a multi-method design across three open datasets, the research examines: (1) firm-level adoption drivers using Eurostat ICT usage data analyzed through a binomial generalized linear model with country fixed effects; (2) implementation performance via a streaming Random Forest experiment on the CSE-CIC-IDS2018 dataset evaluated through ANCOVA; and (3) deployment outcomes through causal inference using propensity-score matching on UK Cyber Security Breaches Survey microdata. Each method directly links adoption determinants, operational performance, and deployment enablers to measurable indicators. The findings reveal that enabling capacities cloud integration, ICT expertise, and structured security training strongly predict ML adoption, while governance quality and throughput constraints define the implementation trade-off between accuracy and latency. Privacy by design mechanisms reduce false positives with minimal performance loss. The study contributes theoretically by bridging adoption and implementation models and provides actionable guidance for enterprises and policymakers to strengthen MLOps governance and sustainable cybersecurity resilience.