The AI Buildout Is Being Questioned From Within

Between May and June 2026, five independent financial-stability institutions flagged AI as a systemic financial risk. The BIS compared the AI investment boom to the dotcom bubble in its June 28 Annual Economic Report ^[1]. The FSB issued two separate warnings: one on June 10 about agentic AI in finance ^[2], and one on May 6 about the $2 trillion private credit market, where AI companies account for over a third of deals ^[3]. The Bank of Canada and the Central Bank of Ireland each cited AI in their May 28 financial-stability reviews ^[4]. The Reserve Bank of India proposed requiring financial institutions to install a "kill switch" on AI models ^[5]. No single authority coordinated them. The BIS, FSB, and RBI each operate from a distinct mandate; the Bank of Canada and Central Bank of Ireland share a mandate type -- domestic financial stability -- but arrived at the same conclusion independently, in different countries ^[4][1]. What matters is the simultaneity: independent institutions, overlapping time frame, same verdict. But the sharper finding is that the contradictions are not just between institutions and markets. They run through both the institutions watching the boom and the companies driving it. Start with a watcher. The Bank of Canada published two assessments of AI in the same month, reaching opposite conclusions:

Then. May 13: Bank of Canada forecasts AI-driven productivity gains, Governor Macklem revises outlook upward; no evidence yet of widespread worker displacement ^[6]

Now. May 28: Same institution warns AI vulnerabilities could trigger a cascading financial-stability crisis ^[4]

One arm of the bank sees AI as a productivity engine; the other sees it as a stability threat. Both can be correct. The institution is arguing with itself about the same technology in the same month. Now the participants. Goldman Sachs CEO David Solomon described his own market in stark terms:

We are definitely in a moment where there's more greed than there is fear. — David M. Solomon

He said this while his firm was underwriting Alphabet's $84.75 billion equity raise -- potentially the largest capital-markets deal in history -- toward AI infrastructure spending projected at $180-190 billion in 2026, potentially rising to $300 billion in 2027 ^[7]. Solomon is describing his market as greed-driven while his firm finances the buildout whose risks he is naming. Goldman Sachs Research separately shifted its preference toward hyperscaler stocks over semiconductor stocks, expecting enterprises to demonstrate tangible AI returns -- an implicit acknowledgment that returns must materialize to justify current valuations ^[8]. Meta posted record Q1 results -- revenue up 33% to $56.31 billion, net income up 61% -- yet its stock fell 9% after Zuckerberg raised capex guidance to $125-145 billion ^[9]. Oracle declined roughly 25% starting June 1 after significant capex commitments ^[10]. Alphabet dipped 4% on dilution concerns from its raise ^[7]. These are not companies being punished by outside critics. They are the AI buildout's biggest spenders, and the market is beginning to distinguish between revenue growth and the spending that funds it. The greed signal extends beyond equities. Bitcoin crashed below $60,000 -- its worst week since the FTX collapse -- as an estimated $400 billion rotated from crypto into AI infrastructure and semiconductor stocks ^[11][12]. Michael Saylor called it a capital rotation rather than a Bitcoin-specific problem ^[11]. BlackRock's Robert Mitchnick said AI momentum is drawing capital away from other markets ^[12]. The counter-evidence is genuine. U.S. GDP grew 2% in Q1 2026, partly attributable to the AI buildout ^[13]. Google Cloud carries a $460 billion backlog ^[14]. Microsoft's AI business reached a $37 billion annual revenue run rate, up 123% year over year ^[15]. AWS revenue rose 28% to $37.6 billion ^[16]. Sundar Pichai said enterprise AI has become Google Cloud's primary growth driver for the first time ^[16]. These are real customers paying real money. But even the revenue success stories carry the risk seed. Alibaba Cloud grew revenue 38%, with AI now 30% of external cloud revenue -- yet net income excluding one-time items plummeted 99.7% ^[17]. CEO Eddie Wu's framing was blunt:

We aim to maintain growth that is faster than the market average in order to gain larger market share and firmly cement our absolute market leadership position... those are the primary objectives, and margin is still secondary. — Eddie Wu

That is the competitive arms race the BIS warned could lead to overinvestment: a capital-intensive push where profitability is deferred indefinitely ^[1]. The leverage data reinforces the concern. Morgan Stanley reported hyperscalers doubled their gross leverage ratio from 0.9x to 1.8x in two quarters ^[18]. UBS found the U.S. led the global credit impulse with $800 billion in additional credit over the past year to fund AI projects ^[18]. Five major hyperscalers are projected to spend over $1 trillion on AI capex across 2025-2026 ^[1]. None of this proves a bust is imminent. The revenue is real, the demand is measurable, and the macroeconomic lift is visible. But the system is processing the risk internally. Five independent institutions arrived at the same warning from different positions. The companies spending the most on AI are being punished for that spending by their own investors. The bank financing the buildout is describing its own market as greed-driven. A central bank is forecasting productivity gains from AI while warning about its cascading risks, two weeks apart. The "AI as growth" consensus is fracturing from within. It just has not fully repriced yet.