AI Credit Crunch: Lenders Rethink Infrastructure Financing

The AI industry's massive infrastructure spending is creating a growing financing gap, as companies pay for GPUs, buildings, and electricity long before revenues cover costs. Lenders are becoming increasingly selective, demanding higher returns and more protection, as evidenced by CoreWeave's bond yielding roughly 15%. The article examines how money flows into AI infrastructure through bonds, private credit, and project vehicles, and highlights key risks including construction delays, aging hardware, and concentrated customers. It uses Meta's joint-venture and equity funding strategies and SpaceX's xAI GPU lease commitments as case studies. The bear case suggests cash flow disappointments could amplify damage across linked suppliers and financiers, while the bull case argues strong sponsors and equity cushions will sustain capital flows. Ultimately, credit decisions are becoming the key determinant of who can build AI infrastructure.