BIS Warns Debt-Fueled $4 Trillion AI Investment Boom Threatens Bitcoin (BTC) Stability

BIS Flags Debt-Driven AI Capex as Stability Risk

The Bank for International Settlements (BIS) has warned that the global artificial-intelligence investment race is increasingly financed by debt and private credit rather than operating cash flow, and that a shortfall in commercial returns could transmit instability across the financial system. In a public address delivered Thursday by BIS General Manager Pablo Hernandez, the Basel-based institution declined to declare an outright bubble, but argued that the optimism priced into AI-linked equities now rests on leverage as much as on earnings. Hernandez stated that capital expenditure at the largest AI firms is growing faster than their internal cash generation, with the gap filled through borrowings and hard-to-value private-credit structures — effectively an investment arms race among technology giants. The official speech text published by the BIS lays out the scale of the cycle: the five largest tech companies plan to spend more than $1 trillion on AI-related projects between 2025 and 2026, while worldwide AI investment is projected to climb from roughly $500 billion today to between $3 trillion and $4 trillion by 2030. Citing Bridgewater Associates research, the address noted that Microsoft, Alphabet, Meta and Amazon alone are expected to deploy some $650 billion in AI infrastructure this year, against a combined market capitalization of about $12 trillion. Hernandez was careful to hedge — he did not say a crash is inevitable — but he insisted that the pace and magnitude of current spending, relative to the commercial returns markets are expecting, warrant genuine vigilance.

Four Historical Manias, One Repeating Script

The BIS chief anchored the warning in history, drawing a direct line from today's spending surge to four earlier technology booms: the 1830s canal mania in Britain, the 1840s railway frenzy that ended in widespread bank failures, the 1920s electrification wave, and the late-1990s FOMO-driven dotcom bubble that wiped out trillions in equity value after 2000. Each episode, he argued, was built on genuinely transformative technology yet absorbed far more capital than final returns could justify — and each subsequent correction spread beyond the affected industry into the broader economy. A second structural concern is the opaque financing web now linking chipmakers such as NVIDIA, AMD and Broadcom to hyperscale cloud providers and AI startups through complex, difficult-to-price contracts. If expected AI profits disappoint, stress would travel backward along that capital chain. Hernandez also flagged the heavy weighting of AI stocks in United States equities: a concentrated reversal would hit household consumption and, given the scale of US shares in global portfolios, propagate to other countries. He added that in some jurisdictions, AI export windfalls are already inflating domestic asset prices. The institution did concede that measurable productivity gains from AI are real — in coding, consulting and professional documentation — but stressed the final economic effect depends on training, infrastructure and competition policy, and that while AI does not change central banks' core mandates, it makes monitoring the economy considerably harder. Readers tracking the market in real time can follow live spot and futures prices on MEXC.

Bitcoin Held Near $77,300 as Risk Reads Deepen

For crypto markets, the address lands in an environment where Bitcoin (BTC) trades near $77,300 as of the latest snapshot, and where risk appetite across perpetual futures venues has historically thinned whenever credit conditions tighten. Our reading of the BIS record is that its most load-bearing point is not the bubble question itself but the transparency gap: opaque private-credit plumbing funding a single thematic trade is precisely the structural fragility that burned investors before, as the collapse of algorithmic stablecoins demonstrated within crypto. If AI-linked capex disappoints, the marginal seller of risk assets is more likely to be a leveraged credit vehicle than a crypto-native fund — making the debt cycle, not token flows, the variable to watch.