The US economy is running on borrowed confidence. Surface-level growth has not erased the warning signs beneath it, and several financial indicators now point toward a recession risk that enthusiasm for artificial intelligence may be hiding.
Warnings published on August 19, 2026, and August 23, 2026, describe an economy supported by AI spending while other measures weaken. Without that spending, the economy would likely already be in a recession. A remarkable achievement, if the goal was to make one sector carry the national economy like an overworked intern.
The European Central Bank has warned that a market correction could follow the current wave of AI hype and spending. The concern is not AI itself, but the money flowing into expectations around it—and the possibility that those expectations have moved far beyond what average annual profits can justify.
Investors are paying significantly more for each dollar of average annual profits. That gap indicates problems in the economy because asset prices are demanding stronger results from businesses at a time when other financial signals are flashing warning lights.
The recession signals are already visible
US Courts reported over 600,000 new bankruptcy filings between June 2025 and June 2026. That represents a 12 percent increase over the previous 12 months, adding a hard measure of financial stress to the broader concerns about markets and economic growth.
Bankruptcy filings do not predict every turn in the economy, but this increase shows that financial pressure is reaching households and businesses. Growth can remain visible in headline numbers while more people and companies struggle to meet their obligations. The two conditions can coexist, which is how economic trouble often manages to look healthy until it does not.
Another warning comes from the relationship between private sector yields and the bank prime rate. When private sector yields rise above the bank prime rate, a recession frequently follows. That signal matters because it reflects a financial system facing higher borrowing costs and greater pressure from lenders and borrowers alike.
The yield curve also indicates an imminent onset of a US recession. Taken together with rising bankruptcy filings and the yield relationship, it suggests that the economy has less room for error than its surface performance implies.
AI is cushioning the decline—not removing it
Financial crisis expert and economist Tuomas Malinen offered the bluntest assessment: “We need to acknowledge that the bottom can fall beneath the US economy, practically, in any minute.” His warning describes an economy that can lose its support quickly when confidence, spending, or market prices turn.
Economist David Rosenberg focused on the role AI spending plays in keeping the numbers afloat. “When you strip out the AI spend, the economy is actually very weak,” he said. That distinction separates growth created by broad economic strength from growth concentrated in a powerful investment theme.
The US economy has not yet dipped into a recession partly because enthusiasm for AI continues to drive spending. Companies and investors are still placing money behind AI-related expectations, giving the economy a source of momentum even as the yield curve, bankruptcy data, and profit valuations point in the opposite direction.
That support cannot be treated as proof that the underlying economy is sound. If AI spending falls, stalls, or fails to produce the profits investors expect, the correction warned about by the European Central Bank could expose the weakness that current enthusiasm is covering.
The central risk is not that AI has no economic value. It is that AI spending has become important enough to delay recognition of broader weakness, while investors pay more for each dollar of average annual profits and financial stress continues to build.
For now, the economy remains outside a recession. The indicators say that status may depend less on durable strength than on how long the AI boom can keep everyone looking elsewhere.
Based on
