Why Europe Must Treat "AI VAT erosion" as a Fiscal Emergency

AI could shrink the VAT base that funds many European public services
Falling wage income means weaker consumption and lower tax revenue
Governments must adapt fiscal policy for the AI economy

In 2023, European governments raised about €1.2 trillion in value-added tax (VAT), accounting for roughly 7.2% of the EU’s GDP and almost 16% of total government revenues. The European Commission reports that VAT remains central to public budgets, but tax revenues as a share of GDP fell in 2023 to their lowest point since 2011. We call this problem AI VAT erosion, meaning a drop in taxable consumption due to lower wage income and changing spending habits caused by automation. This is not any distant worry; it is a real and urgent fiscal issue. The European Commission also notes that the VAT compliance gap in the EU reached 128 billion euros in 2023, showing that shortfalls in funding for key public services may be linked to falling VAT revenues. Policymakers should see this as a revenue problem, not simply a technological trend. Solutions include supporting consumer demand, improving tax collection, and creating fair revenue sources that do not discourage business investment. We have used these tools before, but today’s situation requires action faster and on a larger scale than ever.

Why AI VAT erosion poses an urgent fiscal risk

AI VAT erosion is a serious and urgent fiscal risk. VAT funds essential public services, including education, healthcare, and local infrastructure. Studies using bank transaction data show that when people lose jobs or see their wages fall, they quickly reduce spending, especially on services and non-durable goods, which are subject to VAT. (Targeted policies and household consumption dynamics: Evidence from high-frequency transaction data, 2025) If many people face job losses or lower wages, the overall drop in spending can hurt the national economy. In countries that depend heavily on VAT, even a small two- to three-percent fall in consumption can cause major budget gaps. (Consumption Tax Trends 2024, n.d.) The European Commission reports that tax revenues as a share of GDP in 2023 were the lowest since 2011, putting more pressure on governments to either raise other taxes, cut public services, or borrow more. Local governments often feel these effects first, with budget cuts directly affecting daily life. (Sustainability of Public Finances in OECD Countries, n.d.)

Timing makes this risk even greater. The European Commission reports that tax revenues in the EU-27 have dropped to 39 percent of GDP, the lowest since 2011. This places additional strain on public services, which are already facing reduced revenue from environmental and property taxes. While AI and other technologies may eventually create new jobs and industries, a fall in tax revenue can have lasting effects on local businesses and workforce training if left untreated. (Artificial intelligence’s (AI’s) function in improving tax revenue, institutional quality, and economic growth in selected BRICS-plus countries, 2025) The period between job losses and new job creation is when public support matters most (OECD Economic Surveys: Germany 2025, n.d.). If VAT erosion occurs before new tax bases are in place, key services such as hospitals, schools, and infrastructure could suffer, harming society. (Development, n.d.) The European Commission says policy should focus on both collecting revenue and keeping demand steady, especially since tax revenues as a share of GDP fell in 2023 to their lowest since 2011. Choices made now will decide if automation improves living standards for everyone or increases inequality and hardship. (Rockall et al., 2025) The main goal is to avoid sharp drops in household spending and private investment.

How AI reshapes income distribution and the VAT base

According to a report from the European Commission, artificial intelligence is changing patterns of income distribution and consumption, which has important consequences for the VAT base. These shifts present new tax compliance challenges and affect fiscal sustainability and market competitiveness in the EU (Mind the Gap Report, 2025). Income changes occur in two principal ways. First, automation replaces tasks traditionally performed by workers, cutting labor income in those positions. Second, ownership of AI technologies and data increases corporate profits and rents. According to data from the European Commission, value-added tax (VAT) accounted for 7.2% of the EU's GDP and 15.7% of total government tax revenue in 2023. This means that any shift in income from workers, who typically spend a larger share of their earnings, to those who save and invest more could significantly affect total consumption and, in turn, VAT revenues. According to a report from Le Monde, France is facing a nearly €10 billion shortfall in expected VAT revenue in 2025, showing how declines in consumption can weaken the taxable consumption base even if other parts of the economy, such as profits, grow. This phenomenon is reminiscent of previous waves of technological change that favored capital over labor; what distinguishes the current shift is its rapid pace and broad scope (The transformative power of AI, 2025).

To guarantee transparency, the quantitative assessments presented here rely on publicly available data and conservative assumptions. VAT-to-GDP ratios are based on official EU accounting for 2023, while estimates of AI adoption and exposure draw on recent business surveys and international agency reports spanning 2023 to 2025. In cases where direct measurements of labor displacement are unavailable, moderate scenarios—such as a 3% GDP loss in wage income over five years—are applied to illustrate exposure without claiming exact predictions (European Commission presents Annual Report on Taxation, 2025). Consumption responses utilize detailed marginal propensity to consume (MPC) estimates at the transaction level for unemployed households (Ferreira et al., 2025). The calculations are simple and serve to highlight the order of magnitude of possible impacts. They are not predictive systems and do not incorporate broader economic feedback effects that might increase or reduce outcomes. The intention is to clarify the scale of the issue so that decision-makers can use this information for strategic planning rather than seizing on overly precise forecasts.

Where will fiscal pressure from AI VAT erosion first emerge?

European countries are not all equally at risk from AI VAT erosion. The level of risk depends on how much a country relies on VAT, how many jobs are exposed to automation, and how strong its social safety nets are. Countries with high VAT dependence, large service sectors likely to be automated, and weaker unemployment support will feel the effects sooner (Automation and the employment elasticity of fiscal policy, 2023). A study in Korea shows that automation can hit some communities harder, with job losses among lower-income workers leading to reduced community spending and tax revenues, which can hurt entire towns and cities. The European Commission says national governments may need to step in with fiscal transfers to cover revenue gaps, like the 128 billion EUR VAT compliance gap in the EU in 2023. This is especially important for areas that rely on tourism and small retail, which are more exposed to such financial pressures.

The geographic distribution of risk highlights certain areas of Europe as early warning zones. Many EU member states depend heavily on VAT revenues, and several recent surveys document swift AI adoption in selected industries and regions between 2023 and 2025. According to the European Commission’s latest Data on Taxation Trends report, tax revenues as a share of GDP dropped in 2023 to the lowest level since 2011, suggesting that countries highly dependent on VAT and facing early automation could be at particular risk of fiscal stress. Tourist destinations and small towns that rely on service sector employment may be especially vulnerable. A downturn in regional spending reduces demand for hotels, restaurants, and transportation services, which may lead to permanent business closures that slow recovery and erode the tax base further (Transport, 2025). While these effects begin locally, their fiscal consequences extend nationally, stressing the importance of robust national fiscal buffers and effective local support systems to strengthen resilience (OECD, n.d.).

Governing the age of AI VAT erosion

To confront these challenges, a phased implementation roadmap delivers practical guidance. According to the OECD, implementing VAT reforms and upgrading digital VAT collection systems are important steps for governments aiming to improve tax administration. While the report notes these digitalization efforts, it does not specifically recommend focusing on closing straightforward revenue leaks or providing targeted cash assistance and retraining vouchers in the first year. These initial measures buy critical time and prevent local economies from collapsing. Between the second and fifth years, reforms should aim to broaden the tax base by introducing carefully designed taxes on excess profits, improving property and land taxation frameworks, and establishing cross-border rules for taxing digital revenues. Revenues generated should be invested in lifelong education and local economic regeneration (European Commission presents Annual Report on Taxation, 2025). Over the longer term, strengthening tax administration capacities, building comprehensive public data infrastructures, and advancing international cooperation on taxing digital rents will be essential (Fund, 2025). Each of these steps may seem modest on its own, but collectively they reinforce fiscal robustness while allowing firms to continue investing in AI. The overarching objective is to build robustness rather than impose restrictive measures (Cepparulo & Reitano, n.d.).

The policy response to AI VAT erosion should pursue three main targets: preserving consumer demand, securing efficient tax collection, and expanding a durable tax base. According to the OECD, several countries have responded to shifting economic conditions by increasing their standard VAT rates, demonstrating a broader effort to protect tax revenues and sustain public finances. Temporarily financing these supports through borrowing may be prudent if it prevents a more severe and lasting decline in tax revenues and economic capacity (OECD Sovereign Borrowing Outlook 2023, n.d.). Practical execution demands swift, locally engaged programs that effectively reach affected workers and businesses, along with careful sequencing to align assistance with retraining and job-search incentives (OpenAI has a new program to help SMEs get the most out of AI, 2026).

Securing and modernizing VAT collection is equally important. Revenue shortfalls regularly arise from poor compliance and loopholes in the taxation of digital services (Economics, n.d.). Strengthening invoicing requirements, improving cross-border transaction reporting, and investing in cutting-edge tax-administration technologies to detect evasion can increase revenues without raising tax rates. Real-time or near-real-time reporting systems limit opportunities for tax avoidance (Strengthening tax compliance through implementation of real-time reporting in Malta, 2026). Moreover, improved international data sharing and refined rules for digital platform reporting close gaps that erode the VAT base (OECD/G20 Base Erosion and Profit Shifting Project, n.d.). These reforms bolster fiscal sustainability without altering the fundamental consumption tax structure and improve fairness by ensuring all entities contribute appropriately.

Broadening non-labor revenue sources also deserves careful attention. Debates over wealth taxes, levies on the ultra-rich, and selective capital taxation will intensify. Effective approaches focus on taxing economic rents and excess profits rather than standard returns. Improved taxation of monopoly rents, data-derived rents, and the unimproved value of land can capture unearned gains driven by market power. (Schwerhoff et al., 2022) Joint international schemes to tax returns from digital and data-driven platforms are crucial in this respect (Tax Challenges Arising from the Digitalisation of the Economy – Consolidated Commentary to the Global Anti-Base Erosion Model Rules (2023), n.d.). The aim is to establish a diversified revenue system resilient to declines in wage income. This mix must be equitable, administratively feasible, and compatible with incentives for economic growth (Tax Policy Reforms 2025, n.d.). Political acceptance will require candid communication about how these revenues support public goods and assist workers displaced by automation (Ensuring trustworthy artificial intelligence in the workplace: Countries’ policy action, 2023).

Debating AI taxes, wealth taxes, and fiscal trade-offs

Anticipating criticisms and addressing trade-offs are essential. A frequent worry is that taxing rents or capital could impede innovation. This risk arises primarily when tax policies are blunt or improperly targeted. The solution lies in thoughtful design—targeting economic rents, which represent earnings beyond what is necessary to attract investment, rather than routine returns. Well-focused taxes on excess profits or unimproved land can generate revenue with minimal economic distortion (Nicolay et al., 2023). According to Thiago Dias and his colleagues, consumption taxes are regressive, as the poorest 40 percent of households contribute over half of the total tax revenue for most tax rates. While this is a valid concern, it can be reduced through exemptions for essential goods, targeted credits for low-income households, and strengthened social transfers, which together maintain a stable tax base without increasing inequality. (An extended view on inequality and redistribution in the European Union — The role of indirect taxation and in-kind benefits, 2025, pp. 162-177) Implementing reform will be politically challenging, but the cost of inaction is substantially greater (Tax Policy Reforms 2025, n.d.).

In conclusion, AI VAT erosion represents a clear and pressing fiscal challenge. Addressing it requires focusing on effective policy tools: safeguarding demand where employment displacement is concentrated, modernizing tax collection, and capturing concentrated economic rents without undermining productive investment. This is a narrowly focused, pragmatic agenda that avoids symbolic but ineffective headline-grabbing measures. It treats the issue as a matter of sound public finance and social resilience. Europe has the capacity to adopt AI rapidly while continuing to back essential services such as schools, hospitals, and local communities. Achieving this, however, demands careful planning, swift action on tax collection and social supports, and a long-term shift toward taxing economic rents rather than labor income alone (Tax Policy Reforms 2025, n.d.). The ultimate test is straightforward: can governments maintain vital services while the economy reshapes who earns and who spends? Success in this task will make automation a means for shared prosperity rather than a catalyst for public disinvestment and hardship.

References

Adăscăliței, D. et al., 2024. Weathering the crisis: How job retention schemes preserved employment and incomes during the pandemic. Luxembourg: Eurofound.
Cepparulo, A. & Reitano, V.E., 2025. An assessment of the euro area fiscal stance in 2025 and 2026, considering the flexibility for higher defence spending. Brussels: European Commission Directorate-General for Economic and Financial Affairs.
European Central Bank, 2025. The transformative power of AI. Frankfurt: European Central Bank.
European Commission, 2025a. Annual report on taxation 2025. Brussels: European Commission.
European Commission, 2025b. Mind the Gap Report: Tax gaps supporting competitiveness and fairer tax systems. Brussels: European Commission.
European Commission, 2025c. Strengthening tax compliance through implementation of real-time reporting in Malta. Brussels: European Commission Directorate-General for Structural Reform Support.
European Commission Directorate-General for Mobility and Transport, 2025. Unbalanced tourism: Which European destinations are potentially vulnerable? Brussels: European Commission.
European Parliament, 2023. Nicolay, K., Steinbrenner, D., Woelfing, N. & Spix, J. The effectiveness and distributional consequences of excess profit taxes or windfall taxes in light of the Commission’s recommendation to Member States. Brussels: European Parliament.
Ferreira, M.A., Adelino, M. & Oliveira, M., 2025. The heterogeneous effects of household debt relief. ECB Working Paper Series No. 3034. Frankfurt: European Central Bank.
International Monetary Fund, 2025a. Enhancing tax capacity: Revenue gains from strengthening tax administration. IMF Working Paper No. 2025/219. Washington, DC: International Monetary Fund.
Korinek, A. & Lockwood, B., 2023a. The future of tax policy: A public finance framework for the age of AI. Washington, DC: Brookings Institution.
Korinek, A. & Lockwood, B., 2023b. Public finance in the age of artificial intelligence: A primer. Washington, DC: Brookings Institution.
Milmo, D., 2024. Balance effects of AI with profits tax and green levy, says IMF. The Guardian.
OECD, 2023a. Ensuring trustworthy artificial intelligence in the workplace: Countries’ policy action. OECD Employment Outlook 2023. Paris: OECD Publishing.
OECD, 2023b. OECD sovereign borrowing outlook 2023. Paris: OECD Publishing.
OECD, 2024a. Consumption tax trends 2024. Paris: OECD Publishing.
OECD, 2024b. OECD economic surveys: Romania 2024. Paris: OECD Publishing.
OECD, 2025a. OECD economic surveys: Germany 2025. Paris: OECD Publishing.
OECD, 2025b. Sustainability of public finances in OECD countries. Paris: OECD Publishing.
OECD, 2025c. Tax policy reforms 2025. Paris: OECD Publishing.
OECD/G20 Inclusive Framework on BEPS, 2025. A decade of the BEPS initiative. Paris: OECD Publishing.
Oxford Economics, 2024. VAT gap in the EU: 2024 report. Oxford: Oxford Economics.
Rockall, E.J., Tavares, M.M. & Pizzinelli, C., 2025. AI adoption and inequality. IMF Working Papers. Washington, DC: International Monetary Fund.
Schwerhoff, G., Edenhofer, O. & Fleurbaey, M., 2022. Equity and efficiency effects of land value taxation. IMF Working Papers. Washington, DC: International Monetary Fund.
TechRadar, 2026. OpenAI has a new program to help SMEs get the most out of AI. TechRadar.
Journal of Macroeconomics, 2023. Automation and the employment elasticity of fiscal policy. Journal of Macroeconomics, 75.
Journal of Macroeconomics, 2025. Targeted policies and household consumption dynamics: Evidence from high-frequency transaction data.
Economic Analysis and Policy, 2025. An extended view on inequality and redistribution in the European Union — the role of indirect taxation and in-kind benefits. Economic Analysis and Policy, 87, pp.162–177.
Tax Foundation, 2024. When taxing AI, don’t reinvent the wheel. Washington, DC: Tax Foundation.