The rapid rise of Generative Artificial Intelligence (GenAI) has prompted many firms to engage in their own efforts to develop proprietary GenAI solutions. Using qualitative, inductive field methods, we study how a large German automotive manufacturer navigated different technical, social, and institutional challenges they encountered when they decided to build an in-house GenAI solution. We adopt Simon’s view of design as problem-solving as our lens to inductively conceptualize GenAI development as an iterative process of problem-solution pairing during which both problem and solution spaces co-evolve and shape each other until they gradually align with each other. Our new theory offers a new and processual perspective on the differences between GenAI development and traditional systems development and provides several implications for how firms can better manage future emerging autonomous general-purpose technologies in the future.
