Accountable Agentic Artificial Intelligence in Employee Experience Ecosystems: An Enterprise Information Management Framework

Abstract

ABSTRACT Agentic artificial intelligence (AI) shifts enterprise information management from information support towards autonomous knowledge‐based action. This conceptual paper develops the Accountable Agentic Employee Experience Ecosystem Framework to explain how agentic AI embedded across human resource management, knowledge management, learning, internal communication, employee service, compliance and employee advocacy generates value co‐creation, value no‐creation or value co‐destruction. Integrating socio‐technical systems theory, knowledge management and interactive value formation, the framework positions autonomy scope and knowledge integrity as antecedent conditions; knowledge provenance and traceability, human oversight and formal contestability, and accountability architecture as complementary governance mechanisms; and accountable knowledge autonomy as the central mediating mechanism. Accountable knowledge autonomy supports human agency preservation and shapes interactive value formation in employee–agent encounters, while task criticality conditions the relationship between autonomy scope and accountable knowledge autonomy; knowledge contamination inhibits the development of accountable knowledge autonomy. Value outcomes subsequently feed back into organisational knowledge, governance mechanisms and autonomy boundaries. The paper advances enterprise information management by theorising agentic AI as a governed knowledge‐action system, extends knowledge management from AI‐assisted knowledge generation to autonomous knowledge use, and expands AI in human resource management research to interconnected employee‐facing agentic workflows. It also provides actionable principles for calibrating autonomy, governing knowledge sources, enabling contestability, assigning lifecycle accountability, governing vendors and evaluating operational and relational value.