The ethical dilemmas of AI-driven leadership: analyzing the shift from emotion-based decision-making to data-driven choices

Purpose This study aims to explore the ethical challenges of integrating Artificial Intelligence (AI) into leadership decision-making, focusing on the shift from emotion-based to data-driven approaches. It examines the impact of AI on ethical considerations in leadership, with an emphasis on the role of emotional sensitivity as a moderator. Design/methodology/approach A quantitative approach was employed, utilizing a survey of 300 leaders and managers across various industries in China. The study introduced two novel scales: “AI-supported leadership decisions (AISLDS)” and “emotional sensitivity of leadership decision contexts (ESLDC).” Data was analyzed using exploratory and confirmatory factor analyses, along with regression analysis to test relationships between AI decision-making, emotional sensitivity and ethical considerations. Findings The study found a significant positive relationship between AISLDS and ethical considerations. Emotional sensitivity was found to moderate this relationship, highlighting that AI struggles with ethically consistent decisions in emotionally charged situations, suggesting that human judgment is crucial in these contexts. Practical implications The findings offer guidance for organizations on when AI is suitable for decision-making and when human empathy is essential. It suggests the need for leadership training that emphasizes the integration of AI with human judgment, ensuring that AI complements rather than replaces human compassion and moral reasoning. Additionally, the study informs the development of policies for ethically integrating AI into leadership. Originality/value This study advances AI leadership ethics by demonstrating that emotional sensitivity moderates the relationship between AI-supported decisions and ethical outcomes. Through two newly developed and validated scales (AISLDS and ESLDC), it provides a context-sensitive framework that clarifies when AI can be used ethically and when human empathy is indispensable.