AI Financing: The Fed’s $3 Trillion Challenge and Its Impact on U.S. Competitiveness

From my desk in the Financial District, where the glass towers house the capital that builds the future, the current AI investment surge feels both exhilarating and disorienting. Morgan Stanley’s projection of $3 trillion in global AI infrastructure spending by 2028 isn’t just a number on a screen; it’s the sound of jackhammers at data center sites and the frantic bidding for GPU clusters. But as Fed officials debate how this boom fits into their inflation models, I worry we’re focusing on the wrong dashboard. The immediate challenge isn’t just the price of electricity or skilled labor—it’s understanding the novel and rapidly evolving financial ecosystem funding this $1.5 trillion external financing gap. This isn’t merely a question of economic overheating. It’s a test of whether our monetary policy framework can see past inflation to safeguard financial stability and protect the very process of innovation.

History offers a stark lesson, one I often recall when visiting the Fed’s marble halls just a short walk from my office. In the mid-1990s, with unemployment falling below accepted estimates of its natural rate, pressure mounted for Alan Greenspan to tighten policy. He famously hesitated, suspecting the models were wrong and that a productivity surge was raising the economy’s speed limit. His restraint allowed the dot-com boom—flawed as it was—to fuel investment that reshaped the global economy. The critical counterfactual, as explored in work like Patrick Moran and Albert Queralto’s 2018 paper in the Journal of Monetary Economics, is this: How much of that productivity gain would have been stillborn under a more aggressive rate hike regime? We’ll never know. Inflation from excessive accommodation eventually prints in the data. Productivity lost because investment never happened is invisible, a ghost in the economic machine.

With AI, the stakes are higher and the potential damage more permanent. The infrastructure—data centers, power grids, specialized human capital—creates what economists call cumulative advantages. If this investment cycle is stunted by policy missteps, lowering rates years later won’t magically repatriate it. The expertise, the supply chains, and the ancillary businesses will have coalesced overseas. A nation can recover from a bout of high inflation; catching up from a lost technological epoch is a far taller order. This is the quiet, enduring risk hiding behind the more vocal inflation debate.

Yet focusing solely on the real economy’s capacity constraints misses half the picture. The other, more opaque challenge is the financial architecture blossoming around AI. For decades, the Fed’s dominant models have granted financial variables little independent weight, treating them mostly as leading indicators for inflation or employment. In recent analytical work with Sergey Sarkisyan, I’ve seen how credit spreads and financing costs contain vital signals about distortions that standard gaps miss. This reflects a curious drift in the Fed’s mandate. Created in response to banking panics, its original purpose—financial stability—has been relegated behind inflation and employment in both its models and public discourse. But as the 2008 crisis demonstrated, leverage and interconnectedness can inflict wounds far deeper and more lasting than missing an inflation target by a few ticks.

AI financing makes this oversight particularly dangerous. We’re not looking at a simple bank loan boom. The capital is flowing through venture debt, private credit funds, complex project finance structures, and an ecosystem of intermediaries with layered exposures. The Fed needs a far clearer map of where the leverage, maturity mismatches, and ultimate risks actually reside. What happens if the projected revenues from AI services disappoint, or if today’s cutting-edge $500 million data center faces obsolescence in five years instead of ten? The propagation of losses through this new ecosystem is a black box.

This demands a shift in intellectual capital at the central bank. Post-2008, the Fed became a world expert in banks, housing, and mortgages. That knowledge remains crucial. But the next vulnerability rarely mimics the last. The analytical firepower devoted to dissecting securitized mortgages must now be trained on private markets and novel funding conduits. A central bank exquisitely calibrated to fight the last war is not, by definition, equipped to anticipate the next battle.

The peril lies in using the wrong tool. If the emerging vulnerability is financial—excessive leverage in private AI funding—then reflexively raising interest rates to combat suspected inflation could be catastrophic. It would simultaneously strain the very leverage we don’t understand and raise the cost of the productive investment needed to realize AI’s gains. The outcome could be a financial accident compounded by a generational loss: the forfeiture of U.S. technological leadership as the ecosystem solidifies abroad.

This is not a case for easy money. The Fed has rightly relearned the perils of letting inflation run hot. Nor is it a suggestion that central bankers should pick AI winners and losers. The mandate is subtler. It’s about matching policy instruments to the correct problems and restoring financial stability to its rightful, co-equal place alongside price stability and maximum employment. The goal is to navigate this $3 trillion boom without unnecessarily impairing the capital formation on which America’s long-run prosperity hinges.

The past few years taught the Fed not to underestimate inflation. The coming years will test whether it can avoid being blindsided by complexity. Inflation, eventually, rings the bell. Financial vulnerabilities can fester in silence until they break things. And missing productivity? That loss is silent too, and often permanent. The greatest risk today is not an inflation overshoot. It’s eroding America’s AI advantage before the promised productivity gains ever reach our shores.

• Understanding financial systems around AI
• The importance of capital flow monitoring
• Addressing leverage in private funding
• Matching policy with economic reality
• Restoring financial stability
• Navigating the AI investment boom