Is the AI Industry Heading Toward 'Too Big to Fail'? Fed Officials Warn: Infrastructure Boom Has Surpassed Pre-Financial Crisis Levels

The artificial intelligence (AI) boom continues to sweep the globe, with major tech companies and investment institutions pouring hundreds of billions of dollars into building data centers, purchasing chips, and expanding infrastructure, ushering the AI industry into an unprecedented wave of investment. However, as investment scales rapidly expand, officials at the U.S. Federal Reserve (Fed) have begun closely assessing whether this AI infrastructure boom is gradually accumulating risks to the financial system, potentially echoing past scenarios such as the tech stock bubble or the real estate bubble.

Currently, several Fed officials believe there are no clear signs of an asset bubble forming, nor do they expect a large-scale financial event similar to the 2008 global financial crisis or the 2000 dot-com crash in the short term. Nevertheless, the continuous expansion of AI investment, the uncertainty surrounding returns, increasingly complex financing methods, and rising corporate debt have officially placed AI-related financial risks on the Fed’s monitoring radar.

New York Fed President: Market Enthusiasm High, But AI Not Yet a Bubble

John Williams, President of the Federal Reserve Bank of New York, said in an exclusive interview with Reuters that he does not currently believe the AI market has formed an asset bubble.

Williams pointed out that the defining feature of today’s market is investors’ high enthusiasm and expectations for new technologies, particularly the potential of artificial intelligence, which continues to drive massive capital flows into related sectors.

He explained that investors are trying to answer an extremely difficult question in real time: how much economic value AI can generate and how much return companies will ultimately achieve. Since these questions currently lack clear answers, significant market price volatility is not surprising.

Williams noted that while AI investment has indeed increased corporate borrowing, the primary borrowers at this stage are large enterprises with strong profitability and stable cash flows, meaning current leverage levels do not yet pose a major threat to financial stability.

He stated, 'I’m not particularly concerned about corporate borrowing posing an immediate risk to the financial system.'

AI Infrastructure Investment Soars, Growth Rate Exceeds Pre-Crisis Housing Boom

Torsten Slok, Chief Economist at asset management firm Apollo, pointed out in a research report that while current data center construction has not yet reached the peak of the U.S. housing bubble, its growth rate is highly concerning.

According to Apollo’s analysis, U.S. residential construction investment accounted for approximately 6.6% of GDP at the peak of the housing market in 2005. In contrast, current data center investment remains less than half of that level.

However, when observing the growth rate of investment as a share of GDP, the expansion pace of AI infrastructure has already surpassed the growth rate of the real estate market before the 2008 global financial crisis, indicating that AI capital expenditures are accumulating at an unprecedented speed.

This has led many market participants to worry that if future AI demand growth falls short of expectations, the massive upfront investments could face recovery difficulties, potentially triggering a chain reaction across financial markets.

Fed Officials Worry: Is the AI Industry Moving Toward 'Too Big to Fail'?

In contrast to Williams’ relatively optimistic stance, Jeff Schmid, President of the Federal Reserve Bank of Kansas City, has expressed higher vigilance regarding AI-related financial risks.

Schmid recently stated in a speech that the current financing model of the AI industry has begun to reveal issues worthy of deeper discussion. He believes the market needs to re-examine from a systemic financial perspective whether the AI industry is gradually evolving into another 'Too Big to Fail' sector.

Schmid is particularly concerned about the complex web of capital flows and financial linkages behind AI infrastructure projects.

He noted that data center construction typically involves long-term leases, financing arrangements, energy suppliers, large-scale power investments, local government infrastructure, and community economic development, creating highly interdependent contractual relationships across multiple industries.

He warned that if any single link in this chain fails, risks could rapidly spread through these financial and commercial connections.

Schmid described the situation as a 'circular system' where data centers, energy suppliers, and local communities become highly dependent and heavily indebted—'a single spark could trigger a much broader chain reaction.'

San Francisco Fed: AI Investment Scale Is Concerning, But Some Risks Haven’t Fully Materialized

Mary Daly, President of the Federal Reserve Bank of San Francisco, also stated that simply observing the scale and growth rate of AI investment makes the current situation appear highly concerning.

However, she pointed out that many AI-related investments are still in the stage of corporate announcements, partnership agreements, or planning phases, and have not yet fully translated into completed data centers or fixed assets.

This means that even if market demand cools in the future, some investments can still be adjusted or postponed, so the risk of creating large-scale 'stranded assets' is temporarily not as high as some fear.

'Stranded assets' refer to equipment that a company invests heavily in building but cannot be fully utilized due to insufficient market demand or rapid technological change, leading to premature write-offs and significant asset impairments.

The Fed Establishes AI Risk Monitoring Mechanism to Identify New Financial Threats

Nonetheless, Daly emphasized that the continued use of debt to fund the rapid expansion of AI investment is a critical issue the Fed must monitor.

She stated that the Fed is not merely reviewing past financial crises like the 2008 event, but aims to build a new financial risk monitoring framework to proactively identify novel risks emerging in the AI era.

Daly said the Fed hopes to establish a comprehensive 'risk dashboard' to continuously track AI investment, corporate financing, market leverage, and financial system linkages, identifying potential triggers before localized issues escalate into broader financial risks.