America's $39 trillion national debt is a problem that keeps getting bigger, and everyone from politicians to tech billionaires has an opinion on how to fix it. Now, a new report from the Yale Budget Lab suggests that artificial intelligence might be part of the solution—but only if the government doesn't get in its own way.

The study, released Wednesday, paints a cautiously optimistic picture: moderate AI adoption could boost annual labor productivity growth to 2.5% between 2025 and 2030. That kind of productivity jump could slow the growth of the debt-to-GDP ratio and maybe even start to shrink it. But here's the catch—the government might spend so much money helping workers displaced by AI that it eats up all the benefits.

The report ran two scenarios. In one, the government spends the equivalent of $42,400 per displaced worker (matching retiree spending levels). In the other, it spends $5,500 per unemployed worker (matching current unemployment benefits). In both cases, AI-driven productivity improvements help reduce debt compared to a world without AI. But neither scenario is enough to keep debt from rising beyond current levels.

"It seems unlikely that AI will be some kind of free, infinite money tree," said Martha Gimbel, executive director and co-founder of the Yale Budget Lab, in an interview with Fortune. She emphasized that the size of the productivity shock and the spending response are critical factors.

The report's implication is clear: if the government wants AI to help with the debt, it needs to stop spending heavily on worker support programs. That's a politically charged suggestion, especially since figures like Senator Bernie Sanders and OpenAI CEO Sam Altman have proposed exactly that kind of support. The report argues that policymakers must factor these costs into any discussion of AI's economic benefits.

But worker support isn't the only potential drag. The report also warns that AI-driven automation could shift income from workers to capital owners. Since capital is taxed at lower rates than labor, that shift could shrink federal tax revenues. And faster economic growth from AI might push interest rates higher, increasing the government's debt-servicing costs and offsetting some of the fiscal benefits.

The U.S. national debt has already surpassed the size of the entire economy, hitting 100.2% of GDP by the end of March—the first time since World War II. Total debt now exceeds $39 trillion, and some estimates suggest it could hit $40 trillion by the November elections.

Elon Musk, CEO of Tesla Inc. (TSLA), has previously suggested that technology-fueled growth, particularly in AI and robotics, is the "only way" forward to manage the debt crisis. The Yale report seems to echo Musk's sentiments, albeit with some caveats. Meanwhile, President Donald Trump has aimed to reduce the mounting debt with tariff revenue and massive investments from trading partners.

JPMorgan CEO Jamie Dimon has also weighed in, warning that the debt problem cannot be ignored indefinitely. He cautioned that rising debt could trigger market volatility, higher interest rates, and reduced demand for U.S. Treasuries. Dimon urged policymakers to address the issue, pointing to the unimplemented Obama-era Simpson-Bowles deficit reduction plan as a missed opportunity that could have helped resolve fiscal challenges.

So, can AI save us from the debt crisis? Maybe, but it's not a magic bullet. The Yale report makes it clear that the government's choices—especially around worker support and tax policy—will determine whether AI becomes a fiscal savior or just another line item in the budget.