Change Management for AI in Service Operations

This study develops a multi-level routine dynamics framework explaining how service organizations manage paradoxical tensions arising from artificial intelligence (AI) adoption. AI enhances efficiency and innovation but also creates conflicting demands—such as control versus flexibility and automation versus learning—that challenge conventional change management. Integrating paradox theory, routine dynamics, and dynamic capabilities, the framework identifies four AI-related paradoxes—performing, organizing, learning, and belonging—and links them to a hierarchical system of routines. It illustrates how paradox navigation recurs across individual, routine, and organizational levels through the interplay of ostensive and performative aspects. The study also introduces a paradox-responsive scorecard as a decision-support tool for assessing how effectively organizations handle competing AI outcomes. Overall, it offers a systematic understanding of how service organizations can navigate paradoxes and strengthen decision-making during AI implementation.