Impact of Artificial Intelligence Adoption on Sustainable Project Success: Mediating Role of Project Process Automation

Abstract

The construction and infrastructure sector faces mounting pressure to reconcile escalating project-delivery demands with the imperatives of sustainable development, including economic viability, social equity, and environmental stewardship. Artificial intelligence (AI) adoption has emerged as a potentially transformative lever for addressing these dual challenges; however, the mechanisms through which AI adoption translates into measurable, sustainable project outcomes remain theoretically underspecified and empirically underexplored. This study examines the direct and indirect effects of AI adoption on sustainable project success (SPS) operationalized through the triple bottom line (TBL) framework, with project process automation (PPA) positioned as a mediating mechanism. Grounded in Dynamic Capabilities Theory (DCT), the study conceptualizes AI adoption as a higher-order organizational capability that drives sustainable project outcomes, both directly and by reconfiguring project workflows through automation. A quantitative, cross-sectional research design was employed, with data collected from 250 construction and infrastructure professionals across four global regions using a structured online questionnaire with a five-point Likert scale. Partial Least Squares Structural Equation Modeling (PLS-SEM) in SmartPLS 4 was used to test four hypotheses. The measurement model demonstrated strong psychometric properties, with all outer loadings, composite reliability, average variance extracted, and HTMT values meeting established thresholds. Structural model results confirmed that AI adoption exerts a significant positive direct effect on sustainable project success (β = 0.318, p < 0.001) and on project process automation (β = 0.541, p < 0.001), while process automation significantly predicts sustainable project success (β = 0.387, p < 0.001). Mediation analysis confirmed that PPA partially mediates the relationship between AI adoption and sustainable project success (indirect effect β = 0.209, VAF = 39.7%), supporting all four hypotheses. The structural model explains 44.7% of the variance in sustainable project success. These findings advance the application of DCT to AI-enabled construction project management and offer actionable guidance for practitioners and policymakers seeking to leverage AI investment for sustainable infrastructure delivery.