Paper Type

ERF

Abstract

Organizations are rapidly adopting increasingly agentic artificial intelligence (AI) systems that autonomously generate outputs, route work, and support multi-step tasks. While prior research documents clear performance gains from AI-assisted work, less is known about how these same systems may simultaneously undermine the knowledge environment through employees’ defensive responses. This study develops and proposes a theoretical model explaining how the use of agentic AI can simultaneously increase and reduce work productivity. The model theorizes a direct positive relationship between agentic AI and productivity gains, alongside an indirect negative pathway in which agentic AI heightens job insecurity, which, in turn, increases knowledge hiding and ultimately erodes productivity. To examine these relationships, the study adopts a quantitative, cross-sectional survey design of knowledge workers who use AI-enabled tools or systems and proposes structural equation modeling to test both the direct and indirect paths in the model.

Paper Number

1133

Comments

SIG ADIT

Share

COinS
 
Aug 15th, 12:00 AM

KNOWLEDGE HIDING IN THE AGE OF AGENTIC AI: HOW JOB INSECURITY MEDIATES THE RELATIONSHIP BETWEEN AGENTIC AI AND PRODUCTIVITY

Organizations are rapidly adopting increasingly agentic artificial intelligence (AI) systems that autonomously generate outputs, route work, and support multi-step tasks. While prior research documents clear performance gains from AI-assisted work, less is known about how these same systems may simultaneously undermine the knowledge environment through employees’ defensive responses. This study develops and proposes a theoretical model explaining how the use of agentic AI can simultaneously increase and reduce work productivity. The model theorizes a direct positive relationship between agentic AI and productivity gains, alongside an indirect negative pathway in which agentic AI heightens job insecurity, which, in turn, increases knowledge hiding and ultimately erodes productivity. To examine these relationships, the study adopts a quantitative, cross-sectional survey design of knowledge workers who use AI-enabled tools or systems and proposes structural equation modeling to test both the direct and indirect paths in the model.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.