<p>As artificial intelligence (AI) becomes increasingly embedded in organizational life, human–AI collaboration is emerging as a defining feature of contemporary work. Although research on AI has grown rapidly, the literature in organizational behavior remains fragmented across topics and levels of analysis. To address this problem, we develop a multi-level review framework spanning the individual, organizational, and societal levels to systematically synthesize existing research. At the individual level, prior studies have focused largely on employees’ relatively static attitudes toward AI, while paying less attention to the evolving nature of these attitudes and to emotional dynamics in human–AI interaction. At the organizational level, research shows that AI can enhance productivity, yet it may also intensify job polarization and workplace inequality, raising unresolved tensions between efficiency and employee well-being; moreover, AI use in nonprofit organizations remains underexamined. At the societal level, AI is reshaping labor markets, but it is also associated with employment instability, inadequate social protection, and a widening digital divide. By integrating findings across these three levels, this review clarifies the current state of knowledge, identifies key limitations, and outlines a future research agenda for understanding and responding to AI-driven transformations in work and organizations.</p>