America’s debt problem has become almost too big to visualize at this point.
Gross federal debt now stands at nearly $40.1 trillion, up from about $30.9 trillion at the end of fiscal 2022, which means Washington has added roughly $9 trillion in less than four years. It’s like a household that continues rolling an ever-larger credit-card balance forward.
Nevertheless, veteran economist Mark Zandi believes that AI could offer an unexpectedly powerful way to make that problem more manageable.
That was a key theme on the “Inside Economics” podcast led by the Moody’s Analytics chief economist: Can faster AI-driven productivity actually help America grow its way out of its fiscal hole?
Surprisingly, the answer was more encouraging and lands at a point when AI investors are firmly in “show-me” mode.
Under a scenario resembling the late-1990s productivity boom, the discussion found that stronger economic growth could push the primary deficit close to zero within six or seven years.
But there’s a catch. AI might also create new fiscal costs that eat into much of that benefit.
Zandi says AI could change the debt math quicker than expected
Zandi started from a surprisingly dark place.
The Moody’s Analytics chief economist noted that the federal deficit is running near 6% of GDP, the primary deficit is around 3%, and publicly held debt is roughly equal to the size of the economy.
“I can’t remember a time when all three of those measures were kind of screaming the same thing,” Zandi said. His translation: “We got a problem.”
More AI:
- Nvidia just made a move Wall Street wasn’t ready for
- Microsoft just took sides in AI policy fight
- OpenAI just disclosed something genuinely alarming
The latest CBO baseline tells essentially the same story.
Debt held by the public is projected to rise from 101% of GDP in 2026 to 120% by 2036, while annual net interest costs climb from about $1 trillion to $2.1 trillion.
However, Zandi sees one potential variable the government’s baseline may be underestimating: AI productivity.
He questioned whether stronger AI-driven growth could, “if not bail us out of our fiscal problems, certainly make them a lot easier.”
Brookings economist Ben Harris’ modeling provided the striking answer. If AI helps generate growth resembling the late-1990s boom, the primary deficit could approach zero within six or seven years.
That would not erase America’s existing debt or its interest bill. It would, however, mean Washington largely stops adding new debt before interest costs.
As Harris put it, “Effectively, you stop digging.”
AI’s fiscal rescue comes with several uncomfortable catches
The podcast quickly complicates that optimistic scenario. Higher productivity produces more output and taxable income, but where that growth goes matters almost as much as the amount of growth AI spurs.
Harris modeled multiple offsets, one of which is longevity.
If AI meaningfully improves healthcare and Americans live longer, Social Security and other age-related obligations become more expensive. His scenario increases the annual decline in age-specific mortality from 0.7% to 2%, a potentially meaningful fiscal burden, even though it would obviously be positive for individuals.
Another risk is employment. Harris modeled a three-percentage-point decline in labor-force participation, phased in over four years. He also examined a scenario where labor receives only 30% of new income growth, versus roughly 54% in the CBO baseline, with more gains flowing to capital. Labor income generally generates more tax revenue than capital income.
So far, the labor market is nowhere near that stressful case. U.S. employers added 162,000 jobs in August, according to CNN, while unemployment held at 4.1%, based on the latest BLS report.
A big wrinkle, as the podcast participants noted, is that enormous AI returns could push money toward data centers and compute while leaving non-AI businesses facing higher financing costs.
In other words, AI could make the economy more productive while simultaneously making capital more expensive elsewhere.
For investors, AI productivity now collides with the 5% Treasury
The investor implication is less about whether AI is “good” or “bad” for the deficit and more about whether productivity eventually outruns borrowing costs.
Harris argued that a late-1990s-style growth boom could move the primary budget toward balance. In that scenario, investors might become less worried about Treasury supply, the extra return demanded for holding long-dated government debt could fall, and the 10-year yield could move lower.
Markets are currently moving in precisely the opposite direction. The benchmark 10-year Treasury yield pushed through 5% this week, CNBC noted, reaching levels not seen since 2007 as investors confronted inflation, heavy borrowing, and longer-term fiscal concerns.
Stocks felt the pressure on Sept. 15, Yahoo Finance reported, with the S&P 500 falling 0.45% and the Nasdaq losing 0.78%.
That tension is critical for AI investors. Faster productivity can lift corporate earnings and government tax receipts, but persistently higher Treasury yields also raise companies’ cost of capital and reduce what investors are willing to pay today for earnings far in the future.
That said, the numbers to watch, therefore, are increasingly mundane: productivity growth, tax receipts, labor participation, Treasury yields, and realized returns on AI capital spending.
Related: Elon Musk sends strong signal for SpaceX, Nvidia stocks