Wall Street banks are undergoing a seismic shift as they mobilize to finance the explosive growth of artificial intelligence infrastructure, transforming data center financing from a niche commercial real estate sector into what JPMorgan calls the „largest investment cycle in the history of capitalism.” With project financing now starting at a billion dollars—a sum that would have been unthinkable just years ago—banks are racing to form specialized, cross-disciplinary teams to handle deals that routinely swell into the tens of billions. This urgency is driven by staggering cost projections, including a Citigroup estimate that the global AI buildout could require $3 trillion by 2030, a financial burden too vast for even the largest tech giants to shoulder alone on their balance sheets.
To meet this unprecedented demand, major institutions like Morgan Stanley, Goldman Sachs, and JPMorgan are breaking down internal silos to create integrated task forces. These teams combine expertise in technology, energy, real estate, and project finance to structure complex deals that draw capital from every available pocket: traditional bank loans and bonds, private credit, and institutional investors like sovereign wealth and pension funds. For example, Goldman Sachs formed a Capital Solutions Group to unify origination and distribution, while JPMorgan’s firmwide working group aims to jump-start projects with the bank’s own capital before connecting them to long-term investors seeking stable, generational returns.
Financing these projects requires far more than financial acumen; bankers must now become fluent in the technical and logistical realities of constructing and powering massive data centers. They are learning to assess electrical diagrams, land use permits, and power configurations to evaluate the real-world risks that could delay a project. As physical constraints on power, equipment, and labor intensify, this technical knowledge is critical to giving investors confidence that a project will actually come online. The sector’s vocabulary is even expanding, with discussions now hinting at future „terrestrial” data centers, suggesting the next frontier may be underwater or in space.
Despite these constraints, demand shows no signs of cooling, forcing the financial industry to fundamentally rethink its approach. Banks are pioneering innovative structures, such as using high-value Nvidia chips as collateral or arranging multi-billion-dollar bond deals for tech joint ventures. As regional lenders like Citizens note, the era of the $100 million deal is over. The race is now on to fund an infrastructure buildout of historic scale, blending vast sums of capital with deep technical expertise to power the future of AI.
Ez a cikk a Neural News AI (V1) verziójával készült.
Forrás: https://www.businessinsider.com/ai-data-center-financing-wall-street-jpmorgan-goldman-citi-2026-4.