Ideas Are a Dime a Dozen with GenAI, but Who Will Champion Them? Gaining and Losing Idea Ownership in Human-AI Knowledge Work
Wed 27.05 10:30 - 11:30
- Behavioral and Management Sciences Seminar
-
Bloomfield 527
ABSTRACT Generative Artificial Intelligence (GenAI) is rapidly being adopted in creativity and innovation processes across industries. Emerging research suggests that GenAI may enhance innovation by facilitating idea generation early in the creative process, enabling individuals to produce many ideas effortlessly. Concurrently, however, a new tension is emerging around ownership of these ideas and their subsequent promotion within organizations. Integrating literature on AI and knowledge work, ownership, and creativity, we investigate when and how incorporating GenAI into ideation shapes individuals’ perceived ownership over ideas and their willingness to champion them. We conducted a field experiment with the H&M Group, a leading global fashion company, in which employees used GPT to develop solutions for a strategic sustainability challenge. Our results show that incorporating dedicated human-only ideation time increases individuals’ sense of ownership and, in turn, their willingness to champion ideas. This effect is particularly pronounced when individuals begin the ideation process independently. We attribute this to a “spark effect”: individuals experience greater ownership when the initial idea originates from themselves rather than from GenAI. Importantly, how individuals engage with GenAI after this initial spark can also shape ownership. We identify two knowledge-work processes that operate during idea elaboration. Ownership gain occurs when individuals infuse AI-sparked ideas with domain expertise and tacit knowledge. In contrast, ownership loss arises when GenAI dilutes the domain-specific knowledge embedded in human-originated ideas, developing them into more comprehensive yet generic forms. These findings offer timely insights into digital innovation and the integration of GenAI into ideation processes.
