BitMEX co-founder Arthur Hayes has warned that the debt-fueled artificial intelligence infrastructure boom could culminate in a 2008-style credit crisis, arguing that the resulting government liquidity response could drive Bitcoin (BTC) to $1 million or higher.

In a Tuesday blog post, Hayes contends that investors have mischaracterized spending on data centers and power infrastructure as high-growth technology investment, when in reality it functions more like leveraged real estate. He expects lenders to continue financing excessive construction until a slowdown in AI capital expenditure exposes weaker borrowers.

The thesis links the trillion-dollar expansion of AI infrastructure to a potential new source of liquidity for cryptocurrency markets. Hayes has previously argued that Bitcoin's fixed supply of 21 million coins positions it as a beneficiary when governments expand fiat liquidity to stabilize financial markets — a dynamic he sees echoing the quantitative easing programs that followed the 2008 banking crisis. However, his predicted crisis, government bailout, and subsequent Bitcoin rally remain speculative.

Hayes described the AI boom as a "credit story like 2008 and not an earnings story like 2000." He indicated that BTC could trade between $60,000 and $70,000, with possible downside to $50,000, before the credit cycle and the resulting liquidity response fuel a recovery. Hayes also forecast that Ether (ETH) would reach $5,000 by year-end and said Maelstrom, his family office investment vehicle, intends to build a significant position while selling out-of-the-money ETH put options.

His latest outlook builds on earlier commentary about AI's competing effects on crypto liquidity. On May 13, Hayes argued that US-China competition in AI would encourage bank lending and fiat creation, benefiting Bitcoin (related coverage). On June 4, he disclosed selling HYPE and NEAR after warning that major AI listings could divert capital from crypto (related coverage).

Big Tech Locks In $1 Trillion of Future Leases

The scale of commitments underpinning the AI boom is already visible. On Tuesday, Reuters reported that Microsoft, Meta, Oracle, Amazon, and Alphabet have committed approximately $1.09 trillion to leases that have not yet commenced, primarily for data centers.

Those commitments are nearly four times the roughly $285 billion in lease liabilities already recognized on the companies' balance sheets. Reuters noted, however, that the $1.09 trillion cannot simply be treated as debt because it represents undiscounted payments spread across several years. The commitments nonetheless underscore the magnitude of long-term obligations being undertaken by companies racing to secure computing capacity for AI workloads.

The financial strain across these companies is uneven. According to a separate Reuters analysis, Oracle's debt stood at approximately 4.3 times its earnings before interest, taxes, depreciation, and amortization, while Alphabet, Amazon, Microsoft, and Meta each carried ratios below one.

S&P Global analyst Andrew Chang noted that Oracle's data-center leases, which run for 15 to 19 years, present a key risk because its customer contracts last no more than five years — creating a potential mismatch between long-term obligations and shorter-term revenue.