•  
  •  
 
Communications of the Association for Information Systems

Author ORCID Identifier

Nicolas Bono Rossello: 0000-0003-4624-2753

Anthony Simonofski: 0000-0002-1816-5685

Annick Castiaux: 0000-0001-5290-5993

Abstract

Artificial Intelligence (AI) is increasingly being adopted across the public sector, offering both opportunities and challenges. Despite the abundance of research on AI in government, many studies still highlight a lack of empirical research and evidence-based insights, as well as a need for more theory-driven understanding. In this paper, we extensively analyze this issue by focusing on how AI is conceptualized and, from a sociotechnical perspective, how the literature approaches this topic. To do so, we conduct a narrative literature review that critically examines current research trends across these two dimensions. We identify three traps in the characterization of AI: indefiniteness, homogenization, and arbitrariness. As a result, this body of research is dominated by either single-case studies or broad, conceptual reflections; both of which hinder the development of generalizable knowledge. Furthermore, we find that existing research underrepresents the interplay between social and technical factors, resulting in an overly technocentric narrative. We suggest that these two issues—fragmentation and technocentrism—are related and argue that, to advance the field, future research must better integrate sociotechnical perspectives, focus on developing middle-range theories, and foster more context-sensitive conceptualizations of AI. In this vein, we propose a framework and research agenda to help guide future research.

Download

Share

COinS
 

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