An integrated sociotechnical reference model for aligning artificial intelligence implementation in organizations

<title>Abstract</title> <p>Despite the strategic importance of Artificial Intelligence (AI), organizations struggle to translate ambition into execution. Failures rarely stem from algorithmic shortcomings alone; the deeper issue is misalignment between AI efforts and the broader organizational transformation required to absorb them. Adopting a Design Science Research approach, this study systematically analyzes 45 procedural models, classifying them into five clusters from engineering lifecycles to strategic roadmaps. The analysis reveals disciplinary siloization producing two structural deficiencies: a technocratic bias that sidelines human-centric design, and an internal focus that neglects market responsiveness. To address these gaps, Assembly-Based Method Engineering is applied to construct the Integrated Sociotechnical AI Reference Model (ISAR-M). Embedded decision gates synchronize technical capabilities with organizational readiness across the lifecycle. Expert applicability assessments and scenario-based demonstrations confirm that ISAR-M mitigates temporal asynchrony, offering researchers a structural baseline and practitioners a prescriptive coordination artifact for integrated AI adoption.</p>