According The Information’s article “Why AI Companies are Building Out Wall Street-Style Finance Teams,” AI labs and infrastructure providers are managing massive Capex requirements with the help of experts in structured finance and capital markets.

A recent J.P. Morgan blog states that AI infrastructure financing requires “structural creativity” because timelines are long and cash-flow profiles differ greatly from traditional investment-grade markets. The firm also emphasizes the risks around power availability, supply chain constraints, and permitting timelines, which “can extend project schedules and affect financing structures.”

They know this first hand, as evidenced by their involvement in the $500 billion “Project Stargate” initiative, originating $9.6 billion across two construction loans for the initiative’s Abilene, TX campus. Acting as lead left, sole underwriter and sole structuring agent on both transactions, the engagement illustrates the scale of capital now moving into major AI infrastructure projects.

Also increasingly common are exotic asset-backed securities and Special Purpose Vehicles (SPVs) that structure debt around high-quality credits of AI renters – which allows long-term lenders to provide capital. As stated by Skadden, the vast capital needs of AI means “no one product or market is deep enough to finance all that growth…As a result, developers are increasingly tapping multiple pockets of liquidity across different layers of the capital stack in order to fund their investments.”

In addition, as RCRTech has recently reported, Nvidia’s Wall Street partnerships have fundamentally changed how AI data centers are being built, turning compute into an investable, standardized asset class, with AI chips and servers treated like real estate, power plants, or traditional utility infrastructure. With Nvidia’s $500 billion plan to standardize chip financing, asset-backed capital pools are selling public and private debt in an effort to fund smaller AI buyers. Nvidia is backstopping up to 25% of the risk, a type of circular financing that has Nvidia orchestrating athe money used to buy its own chips. This construct has been criticized for concentrating too much risk across the AI ecosystem, which would become an issue if AI demand or commercial revenue slows down.