Enhancing health-care service delivery through AI-integrated knowledge management systems

Purpose This study aims to investigate how embedding Artificial Intelligence (AI) within Knowledge Management Systems (KMS) affects operational resilience in health care. It examines how AI-KMS integration influences disruption mitigation and dynamic resource reconfiguration, and how these effects are shaped by trust in AI and technical complexity. Design/methodology/approach A structured survey was conducted across 193 public and private hospitals in India. The study uses a three-equation structural modeling approach, estimated via 3SLS, to capture interdependent relationships among AI integration, operational disruption and resource flexibility, while accounting for moderating effects of system complexity and trust. Findings Results indicate that AI-KMS integration significantly reduces operational disruptions and enhances the organization’s ability to reconfigure resources dynamically. However, this disruption-mitigating effect is weakened when technical complexity is high. Conversely, trust in AI strengthens the positive effect of AI-KMS on resource adaptability. Resource reconfiguration, in turn, plays a mediating role in reducing disruptions, reinforcing its strategic value in digital health systems. Practical implications The study offers hospital managers the following actionable strategies: piloting low-complexity AI, embedding AI into protocols, enhancing explainability and monitoring usability. From a society perspective, instead, the findings support patient-centred innovation, more resilient health-care delivery and data-informed policy design in digital health ecosystems. Originality/value This paper bridges the literature on AI-enabled operations and knowledge management by modelling AI as a dynamic, trust-contingent capability. It also contributes a socio-technical contingency perspective to digital transformation in health care, extending existing dynamic capabilities theory to account for the role of trust and complexity in mediating technology outcomes.