AI-DRIVEN PREDICTIVE COST AND SCHEDULE ANALYTICS FOR IMPROVING PERFORMANCE IN SAUDI GIGA-PROJECTS

Saudi giga-projects combine exceptional scope, interdependent delivery packages and evolving resource demands, making cost and schedule control a problem of continuous inference rather than periodic reporting. This critical review examines how artificial intelligence can support earlier recognition of budget exposure, completion risk and feasible recovery options. Thirty peer-reviewed studies published between 2021 and 2025 were selected through a focused search of scholarly publisher and institutional records, with article identities checked against their digital object identifiers. Evidence was organised around baseline estimation, dynamic forecasting, progress sensing, integrated modelling, and implementation governance. Ensemble learning, explainable prediction, continuously updated performance indicators and digital twin data streams show practical promise; however, accuracy reported within individual case studies cannot be treated as transferable performance for Saudi giga-projects. In particular, incompatible work breakdown structures, changing project scope, rare disruption events and weak external validation threaten forecast reliability. A governance-led architecture is therefore proposed that combines package-level data harmonisation, uncertainty-aware cost and time forecasts, causal review of emerging risks, and human-authorised interventions.