From ChatGPT to Autonomous Agents: A Systematic Literature Review and Bibliometric Analysis of Agentic Artificial Intelligence in Organizations

Abstract

The rapid diffusion of generative artificial intelligence has altered the technological and organizational landscape, shifting scholarly and managerial attention from systems primarily designed to generate content toward increasingly autonomous systems capable of planning, reasoning, coordinating actions, and pursuing goals with limited human intervention. This transition has given rise to the concept of agentic artificial intelligence, a broad category encompassing autonomous agents, multi-agent systems, and emerging forms of AI-enabled organizational actors. Despite growing academic and practitioner interest, the literature remains fragmented across computer science, information systems, management, and organizational studies, with limited conceptual integration regarding the implications of agentic AI for organizational design, governance, and leadership. This study addresses this fragmentation through a systematic literature review and bibliometric analysis of the emerging research domain of agentic artificial intelligence in organizations. Following PRISMA guidelines, the study employs a structured search strategy using the Scopus and Web of Science databases and applies performance analysis and science-mapping techniques through Bibliometrix and VOSviewer. The analysis identifies the principal intellectual foundations of the field, the most influential authors, journals, and countries, and the thematic trajectories that have shaped scholarly discussions from early research on autonomous agents and multi-agent systems to contemporary debates on AI governance and organizational transformation. The findings reveal a significant shift from technical investigations of autonomous systems toward questions concerning organizational decision-making, human-agent collaboration, governance mechanisms, and the strategic implications of increasingly autonomous AI systems. Six major research themes emerge from the literature: autonomous decision-making systems, multi-agent collaboration, agentic AI governance, human-agent interaction, organizational transformation and leadership, and ethical and societal risks. Building on these findings, the article develops an Agentic AI Organizational Transformation Framework that conceptualizes agentic AI as a dynamic organizational capability whose outcomes depend on governance arrangements, organizational readiness, and institutional trust. The study contributes to the literature in three ways. First, it provides the first integrative mapping of agentic artificial intelligence research from an organizational perspective. Second, it advances a conceptual framework linking agentic AI capabilities to organizational outcomes and governance mechanisms. Third, it develops a future research agenda aimed at supporting empirical investigations of autonomous AI systems in complex organizational environments. The article concludes that agentic artificial intelligence should not be viewed merely as an incremental extension of generative AI but rather as a potentially transformative organizational phenomenon that may redefine decision-making processes, leadership practices, and the boundaries between human and artificial agency.