Governments built their budgets on payroll taxes. AI could quietly knock that foundation out

Bloomberg has published an examination of a scenario that gets noticeably less attention than AI-driven job losses on their own: what happens to government budgets if artificial intelligence starts mass-replacing the high-paid professionals whose taxes today make up the bulk of state revenue.

Why Income Tax Is the Foundation of Government Budgets

According to the Congressional Budget Office (CBO), the US government will collect roughly $1.8 trillion in payroll taxes in 2026 — compared to just $404 billion in corporate taxes. The picture looks similar in other developed economies: income tax makes up about 46% of Canada's total tax revenue and 52.4% in France. That means ordinary workers' earnings, not corporate profits, remain the primary funding source for governments nearly everywhere.

What the Risk Actually Looks Like

The concern's logic is straightforward: if AI replaces a significant share of high-paid professionals — lawyers, analysts, programmers, accountants — governments could face falling income tax revenue and rising social spending on those who lost their jobs or were forced into lower-paying roles, at the same time. In its 2024 report, the CBO already noted that if employees are permanently displaced by AI or take lower-paying jobs after displacement, tax revenue could decrease.

Not All Analysts Buy the Grim Scenario

There's a notably more optimistic view as well. Some researchers point out that workers displaced by AI are unlikely to permanently exit the labor market — historically, technological shifts have redistributed employment rather than eliminating it wholesale. There's also evidence that AI in practice more often boosts existing workers' productivity than fully replacing them. Under that scenario, economy-wide productivity gains could actually increase tax revenue without raising rates at all — simply by growing the tax base.

Solutions Already Being Discussed

According to Bloomberg, OpenAI has suggested governments consider raising taxes on capital gains, corporate income, or introducing new taxes tied specifically to automated labor — to offset potential revenue shortfalls and fund social programs. Separately, there's discussion of "token taxes" — essentially a consumption tax charged per unit of AI-generated content, which makes economic sense for consumer applications. In summer 2026, representatives from Google DeepMind, RAND Corporation, and the US Federal Reserve System already met at the IMF to discuss AI's potential impact on tax systems, government spending, and GDP.

Why This Matters Even Beyond Policy Circles

Regardless of which scenario turns out closer to reality, the fact that major international institutions are already seriously discussing restructuring tax systems for the AI era is itself a signal: the conversation about mass job displacement by artificial intelligence has moved well beyond talk of individual professions and into a question of government financial sustainability itself.

This material is for informational purposes only and is not investment advice.