Evolutionary dynamics of AI-driven enterprise digital transformation: a cybernetic perspective

Purpose In the AI era, digital transformation has become essential for enterprises to adapt as complex systems to uncertain business environments. In this work, we aim to analyze the evolutionary dynamics of AI-driven enterprise digital transformation (EDT), focusing on its evolution and impact on enterprise performance from a system cybernetics perspective. Design/methodology/approach We employ evolutionary dynamics analysis grounded in game theory to model the EDT process, and perform an econometric and empirical analysis on a sample of Chinese A-share-listed enterprises from 2009 to 2023. The approach encompasses heterogeneity analysis, moderation effect models, such as testing enterprise environmental, social, and governance (ESG) factors and market concentration, and threshold effect analyses. Findings The evolutionary dynamics analysis finds that increasing the risk-averse benefits of traditional enterprises and reducing the additional benefit coefficient would hinder EDT. Increasing the investment of technology enterprises in internal resource restructuring accelerates the transformation process. Empirically, AI significantly contributes to EDT. ESG positively moderates the relationship between AI and EDT, while market concentration exerts a negative moderating effect. Furthermore, AI and EDT show an inverted U-shaped trend when R&D intensity exceeds a specific threshold, while a U-shaped trend is observed when enterprise performance exceeds a specific threshold. Originality/value In this work, we uncover the complex, non-linear cybernetic feedback loops, such as U-shaped and inverted U-shaped, and boundary conditions, such as ESG and market concentration for AI-driven EDT, offering systemic strategies for enterprises adapting to the digital ecosystem.