Algorithmic Ambivalence and Strategic Agility: The Impact of AI Feedback Loops on Organisational Unlearning Processes
Purpose: This paper aims to conceptualise the dual impact of artificial intelligence (AI) feedback loops on organisational unlearning processes and explore how this algorithmic ambivalence shapes strategic agility in technology-driven environments. Design/Methodology/Approach: The paper adopts an integrative conceptual approach, synthesising literature on artificial intelligence capabilities, organisational unlearning, path dependence, and dynamic strategic agility to develop a theoretical framework of algorithmic ambivalence in strategic decision-making. Findings: The study reveals that AI feedback loops operate as double-edged mechanisms. On the one hand, they act as corrective triggers that accelerate the detection of deteriorating routines, foster experimentation, and stimulate organisational unlearning. On the other hand, they can act as reinforcing mechanisms that induce algorithmic lock-in, strengthen cognitive inertia, and foster an illusion of algorithmic objectivity that suppresses critical reflexivity. Organisational unlearning emerges as a vital mediating capability determining whether AI-driven insights translate into strategic agility or algorithmic rigidity. Practical implication: The paper provides strategic guidance for managers to mitigate algorithmic lock-in by fostering a reflexivity-oriented organisational culture, ensuring data source diversity, avoiding uncritical algorithmic authority, and deliberately deploying AI to challenge—rather than merely automate and optimise—established operational routines. Originality/value: This study advances management literature by introducing the concept of algorithmic ambivalence to the organisational unlearning discourse, shifting the prevailing narrative from a purely deterministic view of AI efficiency to a nuanced understanding of how algorithmic feedback loops dynamically enable or constrain strategic agility.