Analysis: AI is unlikely to become the "savior" of U.S. debt; sluggish productivity dividend releases combined with tax base erosion may worsen the debt predicament

BlockBeats News, June 17 — Guolian Minsheng Securities released a research report stating that although the market expects AI-driven productivity improvements to alleviate U.S. debt pressure, based on historical experience and current realities, AI is unlikely to replicate the debt relief miracles seen after World War II or during the Clinton era in the short term. The U.S. debt dilemma remains difficult to overcome in the near future. By the end of 2025, the U.S. national debt will approach $38 trillion, with net interest payments nearing $1 trillion.

The report outlines three paths to reduce debt-to-GDP ratio: lowering interest rates, boosting economic growth, and cutting fiscal deficits. Historically, the U.S. has resolved debt issues in two phases — from 1946 to 1974, relying on post-war high growth and technological transformation, reducing the debt ratio from over 100% to about 20% over 30 years; in the 1990s, driven by the internet revolution and Clinton’s fiscal discipline, achieving an average primary budget surplus of about 3.2% annually from 1996 to 2001.

However, the current debt relief via AI faces two major practical constraints. First, the productivity dividend from AI release has a significant lag. According to the University of Pennsylvania, AI can only boost total factor productivity by 0.05 to 0.1 percentage points from 2026 to 2027, and its contribution will only rise to about 0.2 percentage points by the early 2030s, far insufficient to offset current fiscal pressures. Second, AI accelerates the concentration of factor returns toward capital, systematically eroding the tax base. U.S. personal income tax and payroll tax contribute about 85% of federal revenue, but AI-driven labor displacement and wage suppression will directly impact this main source of revenue; corporate income tax accounts for only about 10% and is taxed at a flat 21%, and with the cross-border tax avoidance capabilities of tech giants, it’s difficult to fill the gap left by individual income taxes, creating a paradox of “technology prosperity leading to a shrinking tax base.”

The report suggests that solutions include increasing capital gains taxes and taxes on the wealthy, levying a “digital element tax” on the profits of large AI models, and exploring a “robot tax” to subsidize workers displaced by technology. However, these face structural challenges such as the difficulty of taxing cross-border AI element flows, the strong political bargaining power of tech giants, and the risk that unilateral tax hikes could stifle innovation. The report concludes that fiscal and tax reforms in the AI era will inevitably be a long-term institutional tug-of-war, and the U.S. debt problem remains a significant obstacle that the U.S. economy will find difficult to overcome in the short term.