The International Monetary Fund has formally elevated AI market concentration to one of three "salient risks" to global financial stability, alongside sovereign debt costs and cyber-attacks. Its twice-yearly Global Financial Stability Report describes these as "plausible near-term catalysts for a sharp tightening of financial conditions and market turmoil." This is not a peripheral concern buried in an appendix — it is a headline finding. The numbers are stark. The "magnificent seven" tech stocks — Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla — now account for more than 35% of the S&P 500, up from 12% a decade ago, with total market capitalisation rising more than 50% in two years. Global investor exposure to US assets, particularly AI, is at an all-time high. As IMF deputy director Athanasios Vamvakidis put it: "If we get a market correction it could have broad implications." What makes this different from a standard bubble warning is the IMF's focus on three specific amplification mechanisms. First, corporate leverage: the AI companies themselves are borrowing heavily to fund infrastructure buildouts. Second, hedge funds and retail investors are layering additional leverage on top, borrowing to speculate in AI stocks. Third — and most distinctive — the IMF identifies $300bn in transactions between hyperscalers and weaker AI firms over just three years, supplemented by $650bn in debt, in exchange for $1.4tn in future financial commitments. This circular financing means a correction in one firm cascades through its counterparties. The Fund is also unusually direct about whether the investment is justified. It notes "growing evidence" that successful AI rollout "may not require investment on the massive scale being undertaken by hyperscaler firms" and that "potential economic gains may not be as large as expected." This is the IMF telling capital markets, politely, that the capex thesis may be wrong. Geographic contagion is asymmetric. Asian markets face the highest secondary exposure after the US. The UK and eurozone are relatively more insulated, though any investor holding US equities would feel the impact directly. The mechanism is wealth effects in reverse: the same portfolio gains that have been supporting global consumption would, in a correction, withdraw that support abruptly. The structural concern is not that AI is worthless — it is that the financial architecture built around AI has the classic features of fragility: concentration in a handful of names, leverage stacked on leverage, and circular obligations that look like revenue until they don't. The IMF warns that in a downturn, "these growing sources of leverage heighten the potential for forced selling and surges in equity volatility." This is the reflexive loop: falling prices trigger margin calls, which trigger more selling, which triggers more margin calls. The report does not predict a crash. It identifies the conditions under which one could propagate globally — and those conditions are currently all present. The question is not whether AI creates value. It is whether the financial instruments layered on top of that value creation have introduced systemic fragility that dwarfs the underlying productivity gains.
