The Impact of Artificial Intelligence Adoption on Project Success in the Oil and Gas Industry: The Mediating Roles of Project Risk Management and Decision Quality

Abstract

Artificial intelligence (AI) adoption has become a strategic priority across the global oil and gas industry, yet the mechanisms through which AI adoption translates into improved project success remain incompletely understood. This paper examines the relationship between AI adoption and project success in oil and gas projects, focusing on the mediating roles of project risk management and decision quality. Drawing on the Theory of Constraints, the Resource-Based View, and Dynamic Capabilities Theory, a dual-mediation conceptual model is proposed in which project risk management and decision quality operate as parallel mediators between AI adoption and project success. A qualitative, secondary-documentary research design was adopted, reflecting the global scope of the research questions and the impracticality of large-scale primary data collection at the level of an individual study. Twenty-four academic and industry sources were purposively sampled and analysed using a six-phase thematic content analysis structured around seven hypotheses. The findings provide qualified support for all seven hypotheses: AI adoption is positively, but conditionally, associated with project success, and this relationship is substantially transmitted through strengthened risk management, particularly in technical, safety, and compliance risk domains, and through improved decision quality, particularly via enhanced information access, predictive accuracy, and decision speed. Evidentiary support for the risk-management pathway is found to be stronger and more direct than for the decision-quality pathway, reflecting the relative maturity of the underlying literatures. The persistent gap between AI investment and realised project benefit documented in the industry literature is best explained not as a failure of AI technology itself, but as a failure to build the complementary organisational capabilities required to convert AI-generated insight into improved project outcomes. Theoretically grounded and practically actionable recommendations, together with priorities for future primary research, are offered.