Artificial intelligence is often discussed as a productivity story: faster growth, smarter firms, and new sources of national advantage. That narrative is incomplete. AI is not just a technological shift—it is a macro-fiscal stress multiplier. If handled well, it can strengthen medium-term prosperity. If handled poorly, it can accelerate fiscal fragility, social instability, and geopolitical risk.
This paper is written for governments facing a simple but uncomfortable reality: existing fiscal frameworks were not designed for an economy where value creation is increasingly intangible, mobile, and detached from labour income. AI compresses timelines, raises capital mobility, and weakens the traditional link between productivity growth and tax receipts. Policy choices made early in this transition matter disproportionately.
The first pressure point is the erosion of labour-based tax systems. AI disproportionately affects professional and high-productivity roles—the very jobs that anchor income tax, social contributions, and fiscal progressivity. Productivity can rise while the labour share of income falls. This breaks a core assumption embedded in most national budget models: that growth reliably translates into labour tax revenue. At the same time, AI-driven wage polarisation thins the effective tax base even if headline tax rates remain unchanged.
Second, public expenditure pressures become structural rather than cyclical. Labour displacement, schooling, retraining, income support, and regional transition costs are not short-term stabilisation measures; they are multi-year fiscal commitments. Adjustment costs arrive early, while productivity gains arrive later—if they arrive at all in affected regions. This timing mismatch increases fiscal stress precisely when political tolerance for consolidation is lowest.
Third, AI sharply increases sensitivity to policy credibility. Capital, intellectual property, and senior human capital are unusually mobile in AI-intensive sectors. Even the expectation of unstable or reactive fiscal policy—such as ad hoc levies, asset taxes, retrospective taxes, or poorly signalled reforms—can suppress investment before a single amount of revenue is raised. In this environment, credibility loss can shrink the future tax base permanently.
The systemic risk emerges when these pressures interact. Declining labour revenues widen fiscal gaps. Fiscal gaps trigger short-term, reactive taxation. Reactive taxation undermines capital confidence. Capital flight reduces the tax base further. Borrowing costs rise. Fiscal space narrows. What begins as a locally rational response to budget pressure becomes a self-reinforcing fiscal derailment loop with international disengagement of investment.
These dynamics do not stop at economics. Rising inequality and uneven AI gains erode trust in institutions, weaken democratic legitimacy, and shorten policy horizons. Governance capacity deteriorates just as medium to long-horizon planning becomes more important. The result is a feedback loop in which fiscal stress, social tension, and political instability co-evolve—raising sovereign risk premia and weakening national resilience.
International competition intensifies the risk. No major economy can slow AI adoption without losing ground. Falling behind the AI frontier carries lasting costs in productivity, tax capacity, and strategic autonomy. In this context, fiscal policy becomes an equilibrium-selection mechanism. Poorly designed responses can shift an economy toward a low-investment, high-instability path that is difficult to escape.
The central message of this paper is straightforward: AI does not merely challenge tax systems—it challenges fiscal statecraft itself. Reactive, short-horizon responses that worked in slower, less mobile economies are now destabilising. The risk is not that governments “fail to tax AI,” but that incoherent fiscal responses undermine expectations, investment, and legitimacy at exactly the wrong moment.
Avoiding this outcome requires predictable multilateral fiscal frameworks, forward-looking investment in resilience and workforce transition, and an explicit recognition that fiscal credibility is now a first-order national security variable. In the AI era, fiscal policy is not just a response tool—it is one of the primary controls determining whether AI becomes a source of durable prosperity or systemic instability.
Thanks for reading The Rethink series. All views expressed are the writer’s own and do not necessarily reflect those of ASRA, its members, or host organization, UN Foundation. Stay connected and keep exploring — discover more Rethink pieces suggested below and under “View” on the Events & News page.
