Hybrid human–AI leadership: exploring the influence of leadership structure on leadership effectiveness and neural activation

Although AI is increasingly assuming autonomous managerial roles within organisations, its leadership effectiveness relative to that of human managers remains lacking. Additionally, AI and humans possess distinct strengths in managerial tasks, suggesting the potential of collaborative human–AI leadership. This study explored the effects of hybrid human–AI leadership on task performance, subjective experience, and neural activation, with the following leadership structures: human only, AI only, human–AI cooperation, and AI supporting human. A total of 80 participants carried out quality inspection tasks led by a human or AI leader under the above structures, with 20 participants randomly assigned to each of the four groups. The results revealed consistent levels of leadership effectiveness between the human-only and AI-only groups, suggesting that AI management could potentially match traditional human management. In addition, the leadership effectiveness was comparable between the human–AI cooperation and AI supporting human groups, indicating the equal viability of these configurations for human–AI collaborative leadership. The human–AI collaborative leadership did not significantly enhance leadership effectiveness compared to the single-leader conditions. The human–AI cooperation group exhibited higher ventrolateral prefrontal cortex activation during AI leader speech, which was likely owing to greater engagement with information processing in the task guidance and performance evaluation.