Michael Burry, the investor known for his bearish call ahead of the 2008 financial crisis, is again warning about the scale of investment behind the artificial intelligence boom. His latest analysis argues that the current capital cycle has reached levels rarely seen in recent decades, raising questions about whether future AI revenues and productivity gains will be sufficient to justify the enormous infrastructure spending.

AI Investment Is Reaching Unusual Levels

Burry points to net capital investment by S&P 500 companies, measured as capital expenditure minus depreciation relative to nominal GDP, which has reached 2.07%. According to his analysis, that is the highest level seen in any previous capital cycle over almost four decades, with the notable exception of the period surrounding the March 2000 Nasdaq peak.

The comparison does not mean the current AI market is identical to the dot-com era. Instead, Burry is highlighting the scale and duration of the investment cycle. During technology booms, companies can continue expanding capital expenditure even after equity markets have reached a peak, potentially creating excess capacity if demand fails to grow as expected.

Hyperscalers Are Driving the Capital Cycle

The largest spending commitments are concentrated among major technology and cloud companies. Amazon, Alphabet, Meta Platforms, Microsoft and Oracle are expected to spend roughly $800 billion on capital expenditures in 2026, with annual spending potentially surpassing $1 trillion in 2027.

Much of this investment is directed toward data centers, advanced computing infrastructure, networking equipment, memory and electricity capacity required to support AI models. The spending is intended to expand computing capacity as demand for AI services grows, but it also increases depreciation and financing requirements.

Burry has separately highlighted the scale of financial commitments surrounding the five companies. He estimates that their combined future lease commitments and commitments to purchase AI-related hardware and equipment amount to approximately $2.7 trillion, contributing to a broader total of about $3 trillion in obligations and other exposures outside traditional balance-sheet measures.

The Key Question Is Whether Returns Can Catch Up

The economic debate is therefore shifting from how much companies are investing to whether the returns generated by that investment will justify the spending. AI could create substantial productivity gains and new revenue streams, potentially allowing companies to generate sufficient cash flow from the infrastructure being built today.

The risk identified by Burry is that spending could continue increasing faster than demand or profitability. If AI adoption slows, computing capacity becomes excessive or technology becomes obsolete more quickly than expected, companies could face higher depreciation, weaker returns on capital and potential asset write-downs.

Markets will be watching AI revenue growth, free cash flow, data-center utilization, capital expenditure plans and financing costs as the investment cycle develops. The critical issue is whether the current spending surge represents infrastructure for a durable productivity transformation or a capital cycle that has expanded faster than underlying returns. That distinction could have implications not only for technology companies but also for the broader S&P 500, corporate credit markets and the U.S. economy.