The Two Economies Inside One GDP Number
The AI investment boom is simultaneously the only thing keeping U.S. GDP growth positive, a driver of the inflation the Fed must fight, and a bubble the Fed cannot risk popping.
Americans are cashing in their credit card points for groceries and gasoline. Over 35% of reward redemptions now go to daily essentials, and gasoline has become the second-most popular redemption category [1]. At USAA, Mike Moran put it plainly.
Prices move faster than earnings, so households need time to readjust. — Mike Moran
The shift is small in dollars but large in what it signals. When households start treating a rewards program like a cost-of-living subsidy, the consumer economy is not thriving. It is coping. The two numbers that define the American economy right now sit in the same GDP report and tell opposite stories. In the first quarter of 2026, AI-related business investment — information-processing equipment for data centers — surged 39.9%. Consumer spending grew 0.5%, the weakest in four years [2]. The second quarter brought a partial recovery: consumer spending rose to 3.2% and AI investment added another 8.4%, but headline GDP still slowed to 1.5% annualized, dragged down by a surge in AI-equipment imports and the Iran conflict pushing gasoline past $4 a gallon [3]. The pattern is consistent. AI capital spending fills the growth gap that consumer weakness keeps opening. Beneath the spending numbers, the consumer base is thinning. The personal saving rate dropped to 2.6% in April, the lowest since June 2022 [4]. A record 111 million adults now sit outside the labor force [5]. The June unemployment rate fell to 4.2% only because 720,000 people stopped looking for work, not because hiring accelerated, and wage growth of 3.5% exactly matches inflation, producing zero real purchasing-power gains [6]. Meanwhile, 9.5 million federal student loan borrowers — one in five — are now in default, with $233.3 billion of the $1.7 trillion portfolio delinquent [7]. Each of these numbers would be a warning on its own. Together they describe households funding consumption by draining savings and running up credit, not by earning more. Three institutions have each identified a piece of this divergence without connecting the whole. The Federal Reserve's June Monetary Policy Report named AI, tariffs, and the Iran conflict as the three primary inflation drivers and warned that growth is "heavily dependent on AI-related capital spending, while housing and consumer spending remain soft" [8]. The WTO reported that a 40% surge in AI-related electronic components trade offset a 9.7% decline in Middle East exports and a 13.6% drop in US merchandise imports, then cautioned that the AI boom is a temporary buffer and that Q2 data will show the full conflict impact [9]. The IMF projected 3% global growth and noted the AI boom "mitigated the economic impact of the Middle East energy crisis," but warned that if productivity gains do not materialize, "frothy equity valuations — particularly in AI-exporting economies" could correct sharply [10]. Each saw one facet: the Fed saw the dependency, the WTO saw the trade offset, the IMF saw the valuation risk. None drew the line connecting them. The banks tell a different story, and it is worth taking seriously. In mid-July, JPMorgan, Bank of America, and Wells Fargo all reported resilient consumer spending, with credit card loans rising 7.3% at JPMorgan to $249.9 billion [11]. Bank of America CEO Brian Moynihan offered the bullish case.
The U.S. economy has proved more durable than expected. — Brian Moynihan
But the same banks noted that lower-income households face the most severe cost pressures, and the resilience they describe is funded by credit expansion and depleted savings, not by wage growth. A consumer who borrows to spend looks durable right up until they do not. The trap is what happens when you connect the three institutional observations into a single picture. AI investment is keeping GDP positive: without the 39.9% Q1 surge in information-processing equipment, the growth numbers would look very different [2][3]. But AI is itself a source of the inflation the Fed is trying to fight. Chip prices have spiked 290%, and software and accessories are up 14.5%, contributing to a CPI of 4.2% alongside the 23.5% energy surge from the Iran conflict [12]. The Fed is now weighing a rate hike to contain that inflation, with Chair Kevin Warsh facing an internal split between hawks who want higher rates and Vice Chair Jefferson and New York Fed President Williams urging patience [13]. The Treasury's own career analysts have already warned of a systemic AI bubble.
We have to stay ahead. — Scott Bessent
A collapse, the draft report found, would send shockwaves through stock markets, chip manufacturers, utilities, and private credit [14]. And the $1.2 trillion tech selloff in late July suggests the fragility is already surfacing. Tesla fell 16%, Alphabet raised capex guidance past $200 billion, and Nvidia dropped 5% [15]. Rate hikes compress the valuations of capex-heavy growth companies: the same companies whose spending is keeping the headline economy afloat. The Fed has diagnosed the disease and identified the cure. The cure would kill the patient's only remaining vital sign.
- 1. Americans Pivot Credit Card Rewards Toward Daily Essentials
- 2. US Economy Grew 2.1% in First Quarter of 2026
- 3. U.S. Economy Slows to 1.5% Growth Amid Iran Conflict
- 4. U.S. GDP Growth Slows as Inflation Hits Three-Year High
- 5. US Labor Force Non-Participation Hits Record 111 Million
- 6. U.S. Labor Force Drops as Unemployment Hits 4.2 Percent
- 7. U.S. Student Loan Defaults Hit Record 9.5 Million Borrowers
- 8. Federal Reserve Cites AI and Tariffs as Inflation Drivers
- 9. WTO Reports AI Boom Offsets Middle East Trade Losses
- 10. IMF Projects Global Growth While Warning of AI Bubble
- 11. Major US Banks Report Resilient Consumer Spending Amid Iran Conflict
- 12. US Inflation Hits 4.2% Amid AI Chip and Energy Surges
- 13. Fed Chair Kevin Warsh Weighs Rate Hike Amid Inflation
- 14. Treasury Draft Report Warns of Systemic AI Market Bubble
- 15. US Tech Stocks Lose $1.2 Trillion in Four-Day Selloff