Leading the Human-AI Paradox: Paradoxical Leadership, Human-AI Complementarity, Responsible AI Agency, and Employee Innovation

Abstract

The increasing use of generative artificial intelligence (GenAI) in organizational work creates a persistent tension between leveraging machine capabilities and preserving human judgment, autonomy, creativity, and accountability. Drawing on paradox theory, this conceptual paper develops a framework explaining how paradoxical leadership can enable organizations to manage this human-AI tension through a both-and approach. The framework proposes two complementary mechanisms. Human-AI complementarity captures the productive integration of distinct human and AI capabilities, whereas responsible AI agency reflects employees' capacity to use AI while retaining critical evaluation, meaningful decision authority, and accountability. The framework also identifies AI literacy as an enabling boundary condition and AI overreliance as a constraining condition that influences whether human-AI collaboration results in augmentation or substitution. By integrating paradoxical leadership, human-AI collaboration, responsible AI, and innovation research, the paper extends paradox theory to AI-enabled work and explains why GenAI may enhance employee innovation under some conditions while generating dependence and diminished human contribution under others. The paper develops 7 propositions to guide future empirical research on leadership in increasingly hybrid human-AI workplaces.