Paper Number

ECIS2026-1722

Paper Type

CRP

Abstract

Explainable AI (XAI) is increasingly prominent in information systems (IS) research, yet prior work notes a lack of theoretical grounding, which hampers theory development and cumulative knowledge building. To address this gap, we systematically review 44 IS studies with human participants, examining explanation scope (local vs. global), dependent constructs (e.g., trust, performance), and the theories (e.g., cognitive fit theory) that inform them. Our analysis reveals a predominance of psychological and cognitive theories, whereas lenses from other fields such as organizational research (e.g., practice-based view) remain underexplored. This paper contributes to theory-driven XAI in IS by providing a theory-centric overview of empirical XAI research, identifying underused but promising theories such as the illusion of explanatory depth, and outlining perspectives for future work on explanation scope, outcome design, and cross-disciplinary theory integration.

Share

COinS
 
Jun 14th, 12:00 AM

Theorizing Explanations: A Review Of Explainable AI In Information Systems Research

Explainable AI (XAI) is increasingly prominent in information systems (IS) research, yet prior work notes a lack of theoretical grounding, which hampers theory development and cumulative knowledge building. To address this gap, we systematically review 44 IS studies with human participants, examining explanation scope (local vs. global), dependent constructs (e.g., trust, performance), and the theories (e.g., cognitive fit theory) that inform them. Our analysis reveals a predominance of psychological and cognitive theories, whereas lenses from other fields such as organizational research (e.g., practice-based view) remain underexplored. This paper contributes to theory-driven XAI in IS by providing a theory-centric overview of empirical XAI research, identifying underused but promising theories such as the illusion of explanatory depth, and outlining perspectives for future work on explanation scope, outcome design, and cross-disciplinary theory integration.

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