Artificial Intelligence Adoption and Managerial Decision-Making: The Role of Human–AI Collaboration and Managerial AI Literacy

Artificial intelligence (AI) is increasingly being incorporated into organizational activities, including information processing, forecasting, customer management, resource allocation, and managerial decision-making. However, the organizational value of AI depends not only on technological adoption but also on how effectively managers collaborate with AI systems and understand their capabilities and limitations. This study examines the relationship between AI adoption and managerial decision-making, with particular attention to human–AI collaboration and managerial AI literacy. The proposed framework considers AI adoption as the independent variable, human–AI collaboration as a mediating mechanism, managerial AI literacy as a moderating variable, and managerial decision-making quality, decision-making speed, and organizational performance as outcome variables. A quantitative cross-sectional research design was adopted. Data were obtained from 200 respondents associated with the real-estate and construction ecosystem, including customers/property buyers, managers, engineers, architects, supervisors, labour personnel, brokers, investors, bankers, finance and accounting personnel, HR professionals, administrative staff, sales and marketing personnel, legal and compliance personnel, drivers, security personnel, cleaners, technicians, and other relevant stakeholders. Responses were measured using a five-point Likert scale. Based on the response distribution supplied for the present study, 180 respondents (90%) selected 5 (Strongly Agree) and 20 respondents (10%) selected 4 (Agree) for each reported questionnaire item. The resulting item-level mean was 4.90, indicating a strongly positive response pattern. The study provides a structured framework for examining how AI adoption, human–AI collaboration, and managerial AI literacy may jointly contribute to managerial decision-making and organizational performance. Respondent-level inferential analysis using PLS-SEM is required before reporting path coefficients, mediation, moderation, significance levels, and explained variance.