The Influence of Emerging Technologies on Risk Management Processes in The UK Retail Industry: A Study of Amazon UK

This study uses the Digital Age Integrated Technology Framework (DAITF) to resolve the paradox that although retail organisations invest heavily in digital risk infrastructure, technology adoption does not lead to benefits in resilience. The study uses a sequential explanatory mixed-methods design and collected data from six senior managers with a mean tenure of 9.8 years and questionnaires from 150 Amazon UK employees with a 64.3% response rate. According to the structural equation modeling, OR was the strongest predictor of adoption, with a beta weight of 0.923, p < .001, f² = 5.76, a large effect size, stronger than external factors (β = 0.907) and investment in DITF (β = 0.724). Most importantly, OR was the DAITF’s main pathway, as it mediated the relationship between external factors – adoption with 33.5% (indirect β = 0.457, 95% BC CI [0.357, 0.561]). The model had excellent explanatory capability with an R² of 0.931. The main cause of deterioration was automation bias, as confirmed with the qualitative analysis with κ = 0.82. All managers saw the main concern with AI as the over-reliance on AI, with one manager saying, “the danger is not that the system is wrong, the danger is that people stop checking.” Purposeful delays in the deployment of the AI systems, ranging from six to nine months, were recognized as an investment with value, and regulatory compliance was recognized as a strategic benefit. This challenges the Technology-Organization-Environment model, showing that external factors affect effectiveness through resilience capabilities.本研究采用数字时代集成技术框架(Digital Age Integrated Technology Framework, DAITF)，旨在解决零售组织虽重金投入数字风险基础设施，但技术采纳却未能转化为韧性收益的悖论。本研究采用序贯解释性混合方法设计，从6名平均任职年限为9.8年的高级管理人员处采集质性数据，并向150名亚马逊英国(Amazon UK)员工发放问卷，有效回收率达64.3%。经结构方程模型分析，运营韧性(Operational Resilience, OR)为技术采纳的最强预测因子，其标准化回归系数β=0.923，p<0.001，效应量f²=5.76，属于大效应水平，优于外部因素(β=0.907)与数字时代集成技术框架投入(β=0.724)。尤为关键的是，运营韧性为数字时代集成技术框架的核心路径，其在外部因素与技术采纳的关系中发挥中介作用，中介效应占比33.5%（间接β=0.457，95%偏差校正自举置信区间[0.357, 0.561]）。该模型具备极佳的解释能力，决定系数R²为0.931。质性分析证实，自动化偏差为绩效恶化的主要诱因，科恩Kappa系数κ=0.82。所有受访管理人员均认为，人工智能(Artificial Intelligence, AI)应用的核心隐患在于对AI的过度依赖，其中一名管理人员表示："the danger is not that the system is wrong, the danger is that people stop checking." 将人工智能系统部署推迟6至9个月的策略被视作具有价值的投资，而监管合规则被视为一项战略收益。本研究对技术-组织-环境模型(Technology-Organization-Environment, TOE)提出了挑战，研究表明外部因素需通过韧性能力方可影响应用成效。