Modes of Boundary Work in Human-AI Collaboration: A Qualitative Meta-Analysis

Abstract

Artificial Intelligence (AI) is increasingly reshaping professional tasks, leading to the decline of some roles and the creation of new ones. However, AI’s position as an “invading actor” within professional systems remains underexplored. Existing research typically treats AI as an exogenous factor in human-AI collaboration, neglecting a crucial aspect—the jurisdictional conflicts arising from AI’s growing influence in professional fields, which challenge the authority of human professionals.These tensions give rise to various forms of boundary work. More importantly, the underlying mechanisms driving this complex relationship remain unclear. To address this gap, we conduct a qualitative meta-analysis synthesizing 42 case studies. Our findings reveal that the introduction of AI into the workplace triggers jurisdictional conflicts, prompting humans to adopt four distinct modes of boundary work: boundary struggling, boundary bridging, boundary retreating, and boundary creating. The tension between professional knowledge and professional practice gives rise to different modes of boundary work. Our research offers a conceptual framework for elucidating the formation of boundary work in human-AI collaboration. This framework incorporates the emergence of AI agency within professional contexts, thereby providing a novel lens through which to examine the development of more complex and dynamic relationships in human-AI collaboration.

Ähnliche Arbeiten