Fed faces $3T AI financing blind spot

The commentary argues that the Fed's focus on inflation and employment may be missing a larger risk: the rapidly evolving financial architecture behind the AI investment boom. Morgan Stanley's projection of nearly $3 trillion in global AI infrastructure investment through 2028, with a $1.5 trillion external financing gap, highlights the scale of this buildout. The authors note that traditional monetary-policy models give financial variables little independent weight, and recent work with Sergey Sarkisyan shows credit spreads contain policy-relevant information about financing distortions that inflation and the output gap miss.

The historical lesson from 2008 is that the Fed failed to appreciate leverage and complexity in the mortgage-finance system until it became systemic. The authors suggest AI is not subprime mortgages, but the institutional lesson is that when financial innovation moves faster than models, understanding where risk accumulates is crucial. They warn that reflexive tightening could expose leverage the Fed does not fully understand while raising the cost of productive investment, potentially leading to a lasting loss of U.S. technological leadership.

The commentary calls for a different allocation of intellectual resources at the Fed, emphasizing that private markets and new funding structures deserve analytical depth comparable to what banks and housing received after 2008. It concludes that missing productivity is harder to detect than inflation, and the risk of eroding America's AI advantage before its gains arrive should not be underestimated.