Too much light blinds: The transparency-resistance paradox in algorithmic management

Gig platforms increasingly harness AI algorithms to manage workers, offering notable efficiency and scalability benefits. However, the rise of worker resistance, such as manipulating algorithms with fake data, poses challenges to these benefits. These algorithms are often perceived as ''black boxes'', leading to issues around transparency. This research thus explores the impact of algorithmic transparency on worker resistance. Using a longitudinal design, we uncovered a paradox: Initially, greater transparency correlates with enhanced fairness perception and reduced resistance. However, beyond a certain threshold, further transparency starts to backfire, leading to decreased fairness perception and amplified resistance. This paradox challenges the prevailing notion that more transparency always leads to positive outcomes. Moreover, we examined the role of human managers, showing that their empathetic support and caring can mitigate worker resistance when transparency fails to foster fairness. This highlights the power of human touch in the algorithm-driven workplace. Overall, these insights suggest a hybrid management model, wherein the cold efficiency of algorithmic managers is complemented by the warm empathy of human managers, offering a blueprint for more productive and harmonious human-machine interactions in the gig economy.