Credit Risks of The Current AI Data Center Infrastructure Investment Boom

The exponential expansion of global AI investment looks set to continue at scale in 2026, with hyperscalers on track to spend $625 billion in aggregate data center capex by year-end. The boom is also lifting players that supply the ecosystem, with near-term demandoutstripping the supply of new data center capacity and corresponding investments materially boosting growth in the broader U.S. economy. Whether AI adoption and monetization will accelerate enough to turn this early investment‑led economic boost into lasting productivity and financial returns is yet to be seen.   But as funding for this expansion shifts from equity markets to debt markets (fueled largely by private credit), some risks have elevated—as a result, in part, from the surge of investments that increasingly rely on debt and the rise of increasingly complex and circular financing structures.   In this episode of the Look Forward Podcast, co-host Molly Mintz explores the credit risks of data center development with Pierre Georges, Managing Director and Head of Infrastructure Research at S&P Global Ratings.   Their conversation explores the 2026 digital infrastructure outlook; analyzes the execution, financial, and contractual risks of the AI-driven data center buildout from a credit perspective; and highlights the development dynamics of power and grid bottlenecks, supply -chain constraints, and tensions between growth, affordability, and decarbonization. We also discuss S&P Global Ratings' view on what could lead to the emergence of winners and losers across the ecosystem, and the leading indicators to watch for an AI investment slowdown—including changes in enterprise adoption and monetization, supply and demand dynamics across the semiconductor supply chain, performance signals from key partners, the scale of hyperscalers' capex plans, and major players' investment behavior.  For more Look Forward content, please visit the Look Forward homepage.