When AI Helps, When It Hurts: A Contextual Research Framework for Integrating Artificial Intelligence into Agile Scrum Workflows

Abstract

As artificial intelligence (AI) increasingly enters agile project management environments, its impact remains inconsistent, boosting efficiency in some cases while disrupting collaboration in others. Rather than assuming AI’s universal benefit, existing literature challenges this assumption and opens a knowledge gap for investigating the organizational and team-level conditions that moderate AI effectiveness. This paper conducts a literature review and proposes a research framework based on the constructs for the contextual analysis of AI in Agile Scrum, exploring when, how, and for whom AI integration enhances or hinders Agile Scrum workflows. It proposes a methodology for future research that is a quantitative-dominant mixed-methods approach, combining semi-structured interviews with a structured survey. The aim of this paper is to identify the contextual factors that shape AI’s impact in Agile Scrum, which will serve as the basis for the research framework.