Hybrid Business Decisions: Does Task Complexity Uplift Delegation to Artificial Intelligence?

Objective: to investigate the effect of the complexity of a hybrid task — hiring people — on the intention to delegate decisions to AI, mediated by motivation for cognitive effort, and to unravel potential pro- and anti-delegation drivers. Theoretical approach: typology of complex tasks and cognitive reasoning literature. Method: we conducted two empirical studies in which task complexity was manipulated on a between-subject basis (low vs. high) to collect perceptive data from 320 U.S. decision-makers. Result: both studies revealed a significant direct and total effect of task complexity on the intention to delegate decisions to AI, but no significant mediating effect. Our thematic analysis showed a preference for humans making the final decision, with AI playing a supportive role, and revealed that, in a hiring context, low complexity leads to self-choices driven by individual factors, whereas high complexity triggers delegation, stimulated by AI’s technical capabilities. Conclusions: our findings represent a theoretical increment in decision sciences, delegation, and task taxonomy by offering insights into how delegation occurs from humans to nonhuman artifacts in a hybrid task context, and by disclosing potential pro- and anti-delegation drivers. We also draw attention to possible over- and under-delegation behaviors that, when combined with other AI-driven decision-making challenges, can undermine gains from AI and impair the attainment of some Sustainable Development Goals, unless properly managed to mitigate the risks.