AI-enhanced demand forecasting: an organizational information processing view

Purpose Demand forecasting is crucial for effective operations and supply chain management, particularly at the manufacturing stage. This study aims to explore the application of Organizational Information Processing Theory in AI-driven demand forecasting, examining how AI reshapes organizational processes and addressing the enablers and challenges of its implementation. Design/methodology/approach Through action research conducted in collaboration with an Italian manufacturing company, this study developed a deep learning-based demand forecasting system. Adopting abductive reasoning, it offers a theoretical examination of the system’s impact, focusing on the interplay between AI implementation and organizational dynamics. Findings This study reveals the organizational changes needed to implement AI in demand forecasting, focusing on iterative adjustments to align information processing amid evolving and uncertain data. Key factors such as data perception, nervousness, and data unreliability affect how information is trusted and used across departments and with suppliers. Building mutual trust and shared interpretive capabilities helps overcome collaboration barriers and reduces risks such as the bullwhip effect, highlighting the importance of negotiation within the organization alongside technology adoption. Originality/value We extend organizational information processing theory by introducing new constructs that capture AI’s implementation challenges, such as data perception and nervousness. Our study shows how AI-specific factors increase information processing demands, while organizational reshipment and trust enhance capacity. This refined framework offers a novel perspective on AI adoption, emphasizing both internal and supply chain information dynamics.