How Organizations Learn and Innovate to Institutionalize Generative AI: Trust, Feedback Loops, and Cultural Friction

Abstract

This study examines how organizations advance in their use of Generative AI (GenAI) tools after initial adoption with particular attention to the dynamics of trust, cultural resistance, and feedback mechanisms. Grounded in organizational learning theory and institutional theory, we propose a framework that captures the factors influencing how firms learn, innovate, and adapt in the post‐adoption phase. Using data from a large‐scale survey of 800 firms currently employing GenAI systems, we apply structural equation modeling to analyze the relationships among trust in AI‐driven decision support, organizational learning orientation, usage maturity, reinforcement processes, and cultural friction. Our findings indicate that higher trust in GenAI contributes positively to a firm's learning disposition, which in turn supports more mature and embedded use of these tools. However, environments characterized by cultural friction weaken this pathway, limiting the effectiveness of learning efforts. These results point to the importance of reinforcing loops and institutional support in normalizing GenAI use, while highlighting cultural resistance as a key barrier to long‐term integration. The study contributes an integrative lens on how learning and institutional forces jointly shape the evolution of GenAI usage within organizations.