Furious pace of AI investment on some Fed officials' radar now

• AI data-center buildout is currently less than half the size of the 2005 housing boom (which peaked at 6.6% of GDP), but investment pace relative to GDP is growing faster than housing did before the financial crisis
• Kansas City Fed's Schmid questioned whether circular financing commitments between data centers, energy providers, and communities are becoming over-leveraged and could create contagion risks
• San Francisco Fed's Daly noted that many AI commitments remain announcements rather than physical realities, reducing 'stranded asset' risk, but emphasized the Fed is building a monitoring dashboard for potential problems

Summary

Federal Reserve officials are increasingly scrutinizing the rapid pace of artificial intelligence investment and its potential financial stability risks, though views differ on the severity of concerns.

Key Officials' Positions:

New York Fed President John Williams downplayed bubble concerns, stating he doesn't see "a bubble kind of situation" despite acknowledging high investor enthusiasm and volatility. He noted that while AI-related borrowing has increased, it's managed by companies with strong earnings, reducing immediate leverage concerns.

Kansas City Fed President Jeff Schmid expressed stronger concerns, questioning whether AI is becoming "another too big to fail" industry. He highlighted worries about complex financing structures and circular commitments between data centers, energy providers, and communities, cautioning that excessive leverage could spark contagion.

San Francisco Fed chief Mary Daly acknowledged the investment growth rate is "very worrisome" but noted that many AI commitments remain announcements rather than physical assets, reducing "stranded asset" risks. She emphasized increased borrowing as a potential concern.

Key Data Point:

Apollo's chief economist Torsten Slok noted the data-center buildout is "less than half the size of the housing boom," which peaked at 6.6% of GDP in 2005, though the investment pace relative to GDP is growing faster than the pre-2008 housing run-up.

Market Implications:

Fed officials are developing monitoring frameworks to track AI financial risks, focusing on leverage, financing linkages, and potential propagation channels. While not anticipating a crisis mirroring the 2008 financial collapse or dot-com crash, regulators are maintaining vigilance over uncertain returns, complex financing structures, and rising debt levels in the unproven AI sector.