The accelerating growth of artificial intelligence, cloud computing, electrification, and digital economies has significantly increased investment requirements for modernising the United States power infrastructure. Rapid expansion of hyperscale data centres, transmission networks, renewable generation, and energy storage systems has intensified competition for capital while exposing limitations in conventional investment appraisal methods that rely primarily on static financial analysis and expert judgement. Consequently, infrastructure investors, utilities, and policymakers require more intelligent, adaptive, and evidence-based approaches for evaluating project bankability and optimising capital allocation under evolving technical, financial, regulatory, and market conditions. This study proposes an AI-driven bankability framework that integrates machine learning, predictive analytics, multi-criteria decision analysis, and financial risk modelling to support strategic investment decisions across data centres, grid expansion, power generation, and energy storage projects. The framework combines diverse infrastructure, operational, financial, and policy datasets to evaluate project viability using objective performance indicators and benchmark outcomes against recognised industry standards. By enabling predictive risk assessment, dynamic portfolio optimisation, and transparent investment prioritisation, the proposed framework enhances decision quality, strengthens infrastructure resilience, improves capital efficiency, and supports scalable deployment of reliable, sustainable, and economically viable power systems capable of meeting future electricity demand.