Artificial Intelligence in Fashion Reverse Logistics : A Feasibility Study of AI-Supported Disposition Decision-Making

Sammanfattning

Background: The rapid growth of e-commerce has significantly increased product returns in the fashion industry, creating operational and environmental challenges in reverse logistics. One of the most critical reverse logistics processes is disposition decision-making, where companies must decide how returned products should be handled to recover value. However, these decisions are often slow, manual, and complex. Recent developments in artificial intelligence have created opportunities to support faster and more data-driven disposition decision-making in returns management. Purpose: The purpose of this study is to explore AI applications that can potentially support disposition decision-making in fashion reverse logistics and to examine under what conditions the adoption of AI is feasible in the fashion industry. The study also investigates other operational challenges within fashion returns management. Method: This study follows a qualitative research approach based on an interpretivist philosophy and abductive reasoning. The research combines a systematic literature review, secondary data analysis, and semi-structured expert interviews. The collected data were analyzed using abductive thematic analysis guided by the Technology-Organization-Environment (TOE) framework. Conclusion: AI can indirectly support disposition decision-making in fashion reverse logistics by facilitating tasks such as return prediction, fraud detection, and product condition assessment. However, successful adoption depends on factors such as data quality, system integration, organizational readiness, and customer expectations. The research consequently suggests that AI adoption in fashion returns management is influenced by TOE conditions.