Multidimensional Cognitive-State Modeling and an Adaptive Cognitive Load Twin for AI-Supported Managerial Decision-Making

Artificial intelligence (AI)-supported decision systems can improve managerial decision-making, but their cognitive implications may depend on how information, explanations, alerts, time pressure, uncertainty, autonomy, and human control are configured. This study develops an Adaptive Cognitive Load Twin using data from 420 logistics and supply-chain professionals in Türkiye who completed four AI-supported managerial decision scenarios, yielding 1680 participant–scenario observations. The framework models information overload, cognitive load, decision fatigue, perceived explainability, trust in AI, automation bias, perceived human control, perceived decision quality, and objective decision accuracy as distinct but related dimensions. Participant-clustered path analysis showed that higher information overload was associated with higher cognitive load, and higher cognitive load was strongly associated with higher decision fatigue; in turn, higher cognitive load and decision fatigue were associated with lower perceived decision quality. Higher perceived explainability was associated with lower cognitive load, greater trust, higher perceived decision quality, and lower automation-bias tendency, while greater perceived human control was associated with higher perceived decision quality and lower automation bias. Higher cognitive load was associated with lower objective decision accuracy. Participant-grouped machine learning models showed that cognitive load was more predictable than decision fatigue or objective accuracy, with information density and time pressure emerging as leading predictive contributors. Model-based simulations indicated that lower information density and progressive disclosure offered the clearest opportunities for adaptive support. The framework extends digital-twin logic toward multidimensional, human-centered monitoring and adaptation of cognitive conditions in AI-assisted managerial work.