Eduardo Porter argued in an analysis for The Guardian that even supercharged economic growth from AI would have a muted effect on US government finances. The White House hopes AI will help the country grow its way out of debt, a position Porter compared to the Reagan-era claim that tax cuts pay for themselves. Treasury Secretary Scott Bessent is counting on AI to help deliver 3% annual growth, a pace rarely reached this century. The argument matters because federal debt has passed 100% of GDP, raising the stakes for any growth plan.
Why 3% growth falls short of fiscal targets
The Committee for a Responsible Federal Budget has calculated what growth would be needed to repair public finances. To reduce the deficit to 3% of GDP by 2036, the economy would need to expand by about 4.4% per year. Balancing the budget entirely would require about 7.2% annual growth. Those thresholds sit well above the 3% target linked to Bessent. The watchdog is described as a fiscal monitoring group, and its figures frame the scale of the challenge.
The expected contribution of AI to growth looks much smaller than those requirements. The Congressional Budget Office estimates that AI will add about 0.1 percentage points per year to growth. That increment does not close the gap between the 3% ambition and the 4.4% pace needed for deficit reduction. It leaves an even wider distance to the 7.2% pace associated with balance. The comparison shows why faster output alone does not resolve the fiscal arithmetic.
Porter also pointed to the composition of income as a constraint on tax collection. He argued that AI would shift income from workers toward owners of capital. In the US, capital is taxed at about half the rate of labour. As a result, the same amount of growth can produce less revenue when a larger share accrues to capital. The structure of taxation therefore weakens the link between higher output and stronger budgets.
What AI spending means for public borrowing
For companies using AI, the analysis suggests treating productivity gains and fiscal relief as separate questions. Faster workflows, automation of routine tasks, and lower unit costs can improve margins without improving federal balances. A small firm may see direct savings from AI tools, while a large enterprise may capture scale effects across operations. Neither outcome changes the tax mix or the deficit path by itself. Investment cases should therefore rest on operating returns, not on expectations of looser fiscal conditions.
The AI buildout also creates direct competition for investor funds. Technology companies are borrowing heavily to construct data centres while US debt exceeds 100% of GDP. Government borrowing and corporate borrowing then draw on the same pool of capital. If private AI projects absorb large sums, financing public debt can become more demanding. That tension is central to Porter's caution about growth-led debt reduction.
The pressure is reinforced by the revenue burden facing large cloud providers. Economists Jared Bernstein and Ryan Cummings estimate that the largest cloud firms need $13.1tn to $18.7tn in additional revenue over a decade to justify current spending. If those targets are missed, Porter wrote, financing US debt could become still harder. The marker to watch is whether cloud revenue and deficit ratios move toward those thresholds by 2036. Failure on the commercial side would leave the fiscal problem unchanged and potentially larger.
