Conflicting impacts of shadow AI usage on knowledge leakage in metaverse-based business models: A Yin-Yang paradox framing

The accelerated evolution of artificial intelligence (AI) technology enables firms to create metaverse-based business models (MBMs) where the use of shadow AI is prevalent. However, employees' use of shadow AI not only provides benefits but also poses risks concerning organizational knowledge leakage (OKL). How to properly manage the usage of shadow AI has become a critically important yet under-researched issue for firms within MBMs. To fill this gap, this paper considers the paradoxical nature of shadow AI usage and thereby adopts a YinYang dialectical frame to investigate the conflicting mechanisms between shadow AI usage and OKL and their impacts on the sustainability of MBMs. Using a questionnaire survey method, we gathered from enterprises within metaverse-related businesses in China and employed moderated hierarchical regression and bootstrap analysis to test the hypotheses. Results show that there existed an inverted U-shaped relationship between shadow AI usage and OKL and a negative link between shadow AI usage and the sustainability of MBMs. We also examined the mediating role of OKL and the moderating role of AI knowledge-oriented human resource management practices. Theoretically, this article conducts an unorthodox Yin-Yang paradox frame to address the dynamic tensions among shadow AI usage, OKL, and the sustainability of MBMs, enriching the theoretical underpinning of cross-disciplinary research at the intersection of technological change, knowledge management, and business sustainability. Practically, our findings allow practitioners and policymakers to gain a better understanding of the paradoxical interplay between employees' unethical behavior toward AI usage and its consequences. This indicates the unconventional yet crucial challenges facing firms within MBMs in the process of fulfilling the United Nation's Sustainable Development Goals.