ºÚÁϳԹÏÍø

Abstract:Organizations increasingly rely on artificial intelligence (AI) to improve both the quantity and quality of innovation. However, while a growing body of research suggests that AI is well-equipped to improve innovation-related search processes amid distributed information, several recent studies offer cautionary evidence that organizations often fail to benefit from AI-produced findings. In this study, we theorize the cross-cutting effects of AI on the realization of serendipitous innovation, wherein surprising findings give rise to updated or emergent innovation goals and action. Specifically, we propose that serendipitous innovation requires exposure to new and unexpected information but also that organizational actors collectively imbue that information with significant meaning, such that action is undertaken despite the initial lack of fit with existing cognitive schemas. We draw from theory on innovation search and sensemaking to theorize why despite its significant advantages for improving search processes AI can challenge the attribution of meaning in the context of surprising information. Moreover, we highlight the technological and organizational design responses that are necessary to overcome these challenges, thereby producing AI-human interactions that allow for accelerating serendipitous innovation. Our theory contributes to emerging research on the role of AI in innovation, while theoretically bridging the innovation search and sensemaking literatures.