BlackRock CEO Warns AI Data Center Freeze Could Trigger US Recession

BlackRock CEO Larry Fink has warned that halting artificial intelligence data center construction could push the United States into recession, highlighting how dependent parts of the economy have become on massive investments in computing infrastructure.

Speaking to CNBC on October 14, Fink argued that a construction moratorium would threaten economic activity beyond the technology sector, including manufacturing, employment and energy infrastructure. His warning comes as investors grow increasingly concerned about the sustainability of AI spending and the debt financing its expansion.

"If we have a moratorium of some sort in this country, we will be in a recession," Fink said during his CNBC interview.

The warning reflects the extraordinary scale of the AI infrastructure boom, which has created demand for everything from specialized semiconductors to electrical grids and construction workers.

Building modern AI facilities requires enormous quantities of equipment, electricity and skilled labor. Electricians, welders, plumbers and equipment manufacturers all benefit from projects that can take years to complete.

A sudden construction freeze could therefore reduce capital investment across several industries simultaneously, potentially weakening employment and economic growth.

Estimates cited in coverage of Fink's remarks suggest annual AI capital expenditure could exceed $1 trillion in 2027, although actual spending will depend on financing conditions and the pace of project development.

Fink's warning comes as economists increasingly examine the role of artificial intelligence spending in supporting American economic growth.

In its October economic outlook, the International Monetary Fund estimated that a substantial AI investment correction could leave US GDP approximately 1.8% below its baseline projection by 2028.

The figure does not represent a forecast of an outright 1.8% economic contraction. Instead, it measures the potential loss of economic output relative to a scenario in which investment develops as expected.

The risks become more complicated when considering the financing structures supporting major AI projects, where developers often depend on future customer contracts to secure loans before their facilities begin generating substantial revenue.

If financing becomes more expensive, construction could slow even without government intervention. Companies may postpone projects when expected computing revenue no longer justifies the cost of borrowing.

These pressures are already visible in the bond market, where CoreWeave debt has faced significant selling pressure as yields approached 15%.

The company's default insurance costs have also climbed above 1,000 basis points, reflecting heightened investor concern about the financial risks of heavily indebted AI infrastructure operators.

BlackRock has significant commercial interests in private infrastructure investment, making Fink's position particularly relevant as the industry faces questions about its financial sustainability.

The asset manager reported $15.49 trillion in assets under management during the third quarter, alongside approximately $205 billion in net client inflows.

Despite financing concerns, demand for advanced semiconductors remains substantial. TSMC's quarterly profit increased approximately 69%, illustrating how AI chip manufacturers continue benefiting from infrastructure expansion even as investors question valuations.

The contrast highlights a growing economic dilemma. AI spending can support employment and corporate earnings, but excessive borrowing could make the expansion increasingly vulnerable to changes in interest rates and investor sentiment.

For now, Fink's warning underscores the potential economic consequences of abruptly stopping construction. Whether AI infrastructure can sustain its contribution to US growth will depend on companies generating sufficient returns to support continued investment.