Delegating to the algorithm: how AI-driven leadership shapes bias perceptions, trust and emotional disengagement in the workplace

Purpose This paper investigates employee attitudes and behaviors towards the delegation of leadership to artificial intelligence (AI). This work identifies perceived algorithmic bias (PAB) as the psychological construct that links the delegation of leadership to employee disengagement (ED) and biases towards algorithmic trust (AT). In addition, this work identifies PAB as a potential moderator and further analyzes the construct within the context of self-determination theory and the theory of planned behavior. This work also analyzes the PAB and ED of employees within the digital work environment. Design/methodology/approach Data were obtained in a two-wave survey involving 241 respondents employed in AI-saturated technology, logistics, government services and digital finance in the UAE. Wave 1 measured ALD, PAB and AL, while Wave 2 measured ED and AT. Structural equation modeling (SEM) was used to measure the relationships, direct, indirect and moderated, between the variables of the study. Findings The results demonstrate ALD's beneficial impact on ED and PAB. There are significant mediations of PAB on the ALD and outcome variables, from the employees' perspectives on biased trust and disengagement. Further, AL moderates the impact of the PAB on the ALD, whereby higher algorithmic literacy weakens the biases perceived. Originality/value The study contributes to the growing body of literature on algorithmic management by intertwining motivational and behavioral theories to clarify employees' reactions to AI-driven leadership. It shows how digital proficiency and fairness appraisals interact to determine the quality of workplace experiences in environments governed by automated decision-making.