Agenticness of Artificial Intelligence in the Fashion Industry : A Case Study Based on a Conceptual Expansion Structure

This study reconceptualizes “agentic artificial intelligence (AI)” in the fashion industry, shifting the focal point from individual performance metrics to a multidimensional structure of agency. Employing a qualitative case study methodology, the research analyzes 11 AI-driven fashion services through an analytical framework composed of five functional dimensions (autonomy, decision-making, problem-solving, interactivity, and planning/execution)mapped across three operational levels (behavior, structure, and system). Findings indicate that fashion AI does not converge on a single developmental stage but forms a multilayered ecosystem where agency is selected and expanded based on strategic goals and service contexts. The services were categorized into three distinct levels: (1) Conventional AI Agent: Represented by H&M Chatbot, focusing on behavioral automation and stable rule-based recommendations, (2) Potentially Agentic AI: Including Alta, AskRalph, Heuritech, MaIA, OneOff, and WGSN Fashion Buying, which expand into structural decision-making but remain in a transitional phase dependent on human oversight, and (3) Highly Agentic AI: Comprising Shopify, Gensmo, Daydream, and Stitch Fix, which implement system-level agency by orchestrating complex “agentic commerce” processes such as multimodal interpretation and multistep execution. These results demonstrate that AI in fashion is evolving beyond mere assistants toward becoming independent“agentic actors.” This study provides a diagnostic framework and a phased roadmap—from operational efficiency to organizational reorganization—to support fashion companies in their strategic AI integration.