How Do Factors Shape Human Integration of AI-Generated Knowledge?
Baoliang Hu, Jie Mei, Keyi Fang, Shuai Yan · Academy of Management Proceedings · 2025
The integration of artificial intelligence (AI)-generated knowledge into innovation has become a critical issue of both theoretical and practical significance. This study employs structural equation modeling and the Bootstrap method to explore the factors influencing human integration of AI-generated knowledge. The findings reveal that this process aligns with the judge–advisor system paradigm, with influencing factors identified across four dimensions: the human judge role, decision task characteristics, AI advisor role, and AI-generated knowledge characteristics. Specifically, factors such as AI operational capability, knowledge distance between humans and AI, knowledge complementarity between humans and AI, human cognitive load regarding AI-generated knowledge, trust in AI, and decision task complexity are identified as key determinants. Additionally, cognitive factors, including cognitive load and trust, are found to be fundamental, as they not only directly influence the integration process but also mediate the effects of other factors. These findings provide valuable insights into optimizing human–AI collaboration in innovation.