WHEN AI TAKES INITIATIVE: MANAGERIAL AUTHORITY, ACCOUNTABILITY, AND DECISION QUALITY IN AGENTIC AI SYSTEMS

Abstract

The deployment of agentic artificial intelligence systems reached a critical inflection point in 2025 as organizations increasingly allowed AI systems to initiate actions, make recommendations, and execute decisions with limited human intervention. While such systems promise efficiency, scalability, and enhanced decision quality, they also raise fundamental questions regarding managerial authority, accountability, and governance. This study examines how the growing autonomy of agentic AI systems reshapes managerial roles and decision-making structures in organizations during 2025. The research analyzes shifts in authority allocation, accountability mechanisms, and perceived decision quality arising from AI-initiated actions. By integrating managerial, ethical, and operational perspectives, the study contributes to understanding how organizations can harness agentic AI while maintaining responsible oversight and strategic control.