Analysis: AI Unlikely to Become a "Savior" for US Treasuries; Slow Productivity Gains and Tax Base Erosion May Further Aggravate the Debt Predicament

BlockBeats news, on June 17, Guolian Minsheng Securities released a research report pointing out that although the market hopes AI-driven productivity improvements will alleviate U.S. debt pressure, both historical experience and current reality indicate that, in the short term, AI is unlikely to replicate the post-World War II or Clinton-era debt reduction miracles. The U.S. debt predicament remains difficult to overcome in the near term. By the end of 2025, the scale of U.S. national debt will approach $38 trillion, with net interest payments nearing $1 trillion.

The report outlines three paths to lowering the debt ratio: reducing interest rates, boosting economic growth, and compressing the fiscal deficit. Historically, the United States managed to reduce its debt ratio in two phases—between 1946 and 1974, relying on post-war high growth and technological transformation, the debt ratio fell from over 100% to about 20% over 30 years; in the 1990s, thanks to the internet revolution and the Clinton administration’s fiscal discipline, the country achieved an average primary budget surplus of approximately 3.2% annually from 1996 to 2001.

However, the current AI-led debt reduction is facing two major practical constraints. First, there is a significant time lag in realizing AI’s productivity dividend. According to estimates from the University of Pennsylvania, between 2026 and 2027 AI can only boost total factor productivity by 0.05 to 0.1 percentage points, and its contribution will only reach about 0.2 percentage points by the early 2030s—far from sufficient to offset current fiscal pressures. Second, AI accelerates the concentration of factor returns to capital, systematically eroding the tax base. Personal income and payroll taxes together contribute about 85% of U.S. federal revenue. The labor substitution and wage compression driven by AI will directly impact this main tax source. Corporate income tax accounts for only about 10% and is subject to a uniform 21% rate, and, combined with the tax avoidance capabilities of tech giants, it’s hard to fill the individual tax gap. This creates the paradox of “the more prosperous the technology, the more depleted the tax base.”

The report suggests that possible solutions include raising capital gains and high-earner tax rates, levying a “digital element tax” on major AI model commercial revenues, and exploring a “robot tax” to subsidize those affected by technological unemployment. However, these measures face structural difficulties including cross-border mobility of AI factors hindering tax administration, the strong political bargaining power of tech giants, and the risk that unilateral tax hikes may suppress innovation. The report concludes that fiscal and tax adjustments in the AI era are destined to be a long-term institutional tug-of-war; the U.S. debt issue remains a major barrier that the American economy is unlikely to overcome in the short term.