An AI-led improvement in worker productivity could help major economies manage their public finances more comfortably, economists told Reuters, but it is unlikely to single-handedly reverse mounting sovereign debt pressures. Nations already face debt-to-GDP ratios above 100% across much of the developed world, and demographic shifts, rising interest costs and new spending priorities for defence and climate policy point to higher liabilities ahead.
Policy-makers in the United States have expressed optimism that AI-driven growth can lift productivity after a lengthy post-2008 slowdown. Economists interviewed for this analysis said the technology has scope to make workers more efficient and reallocate labour toward higher-value activities. Greater output would, in principle, make it easier for governments to manage deficits and roll over debt without triggering market alarm.
To gauge the potential impact of sustained AI-driven productivity on public finances, the Organisation for Economic Co-operation and Development and three economists provided preliminary estimates. Filiz Unsal, deputy director of economic policy and research at the OECD, said an AI productivity surge that raised employment could reduce projected sovereign debt across OECD countries by about 10 percentage points from the roughly 150% of output the OECD expects in 2036. Under that scenario, debt would still be markedly higher than present levels - near 110% today.
How much fiscal breathing room AI can create depends on several interlocking outcomes: whether AI ultimately produces net job gains rather than losses; whether firms shared productivity gains with workers through higher wages; and how governments control spending as revenues change. In the United States, two of the economists projected a slower rise in debt - to roughly 120% of output over the next decade from near 100% today - under an optimistic scenario. A third economist saw little change in U.S. debt from AI-driven gains.
"Productivity is like magic... It helps the fiscal dynamics dramatically," said Idanna Appio, formerly at the New York Federal Reserve and now a fund manager at First Eagle Investment Management. "But our fiscal problems are well beyond what productivity can fix."
Despite the upside that better productivity may deliver, several prominent forecasters remain cautious. Ratings agency S&P, for example, assumes no significant public finance effect from AI by the end of the decade. Mark Patrick, head of macro and country risk at Teachers Insurance and Annuity Association of America, summed up the risk: the best-case path many hope for - essentially being "saved by the bell" - is not something markets can rely on with a clock set against growing obligations.
Cross-country variation is likely. OECD research suggests Britain could capture productivity gains on a scale similar to the United States, while economies such as Italy and Japan might experience roughly half the improvement because of lower AI adoption and smaller shares of sectors that could benefit significantly from automation. Ultimately, the net effect on debt will be a function of fiscal policy choices and demographic pressure.
Demographics looms as the central constraint. Kevin Khang, head of global economic research at Vanguard, emphasised that ageing populations and entitlement spending tied to pensions and healthcare are the root causes of long-term fiscal strain. He said putting public finances on a sustainable path requires broader fiscal repair, with AI providing only temporal relief.
Khang described a scenario he deems most likely in which AI lifts U.S. growth to an average of 3% through 2040. The Federal Reserve’s view of potential growth is nearer 2%. Under Khang’s higher-growth projection, rising tax receipts would slow the expansion of U.S. debt to about 120% of output by the late 2030s. By contrast, if AI fails to deliver, growth stalls and borrowing costs climb, he estimated U.S. debt could reach as high as 180% of output.
Bond markets have shown sensitivity to fiscal slippage since yields rose sharply after the pandemic. That readiness to punish perceived fiscal laxity amplifies the stakes for governments attempting to rely on productivity growth alone. Appio pointed out another demographic headwind for the U.S.: falling immigration, which reduces labour supply and can offset gains from productivity. She added that without AI, the outlook would be notably worse, underscoring AI’s relative importance even if insufficient by itself.
How productivity gains translate into public revenues is itself uncertain. Economy-wide productivity increases should raise tax receipts through higher output, but the composition of gains matters. If AI produces fewer jobs and concentrates returns in corporate profits and capital - income that is often taxed at lower rates than labour - fiscal receipts could disappoint. On the expenditure side, efficiency gains in government operations could lower costs, but there is a countervailing risk that public spending grows alongside a larger economy.
Kent Smetters, director of the Penn Wharton Budget Model analysis group at the University of Pennsylvania, expects only a modest effect on U.S. debt over the next decade from AI-driven productivity. Even if growth exceeds his baseline, he said, that would do little to curb entitlement spending. Social Security alone accounts for roughly one fifth of federal spending, and benefits are indexed to average wages, muting the impact of higher output on the program’s fiscal burden. Other government liabilities tied to labour costs would similarly rise if wages increase.
Filiz Unsal at the OECD noted the importance of wage dynamics: wage growth would be more likely if AI did not expand employment, she said, implying that the distribution between wages and employment is central to fiscal outcomes. Economists also highlighted that changes in productivity could influence real interest rates - a factor that would affect debt service costs - but whether productivity pushes rates higher and for how long growth outpaces any rate rise remains an open question at central banks including the Fed.
All analysts emphasised the high uncertainty enveloping these projections. A sudden economic downturn could truncate an AI-driven upswing, allowing market nervousness about fiscal trajectories to surface before productivity gains materialise. Christian Keller, Barclays’ global head of economic research, warned that a recession could mean "the AI boom may not come quick enough before the market gets nervous about the fiscal trajectory."
For fiscal authorities, the consensus view among the economists is straightforward: AI can alter the tempo of public finances and ease pressure in favourable scenarios, but it does not replace the need for structural fiscal adjustments. Whether AI ultimately expands employment, how widely wage growth spreads, and the policy choices that governments make about taxation and spending will determine if the technology is a temporary reprieve or a more durable fiscal accelerator.