Entrepreneurial Leadership Reconfiguration Under Algorithmic Anchoring Effects: A Mixed-Methods Study

Generative AI agents are embedding into entrepreneurial decision-making, shifting leadership from “human-only” to “human-AI symbiosis.” Existing research largely focuses on AI’s “empowerment” effects, yet lacks systematic understanding of its potential cognitive dependence risks. This study aims to challenge the linear empowerment assumption of AI by proposing and validating a recursive cognitive reshaping model.

Using a sequential mixed-methods design, we find that AI usage intensity positively predicts “algorithmic anchoring,” a relationship negatively moderated by decision sovereignty awareness. Algorithmic anchoring exhibits an inverted-U relationship with entrepreneurial performance, with leadership reconfiguration playing a partial mediating role. High decision sovereignty awareness shifts the inverted-U curve rightward and flattens its descending slope.