A Multi-Jurisdictional Evaluation of Government Revenue Loss Under Full AI Adoption (2020–2026 and Beyond)

Primary Case Study: Federal Republic of Nigeria

With Cross-Continental Comparisons: United States • Germany • China • Brazil • Australia • Japan • South Africa • India • Kenya

EXECUTIVE SUMMARY

The accelerating adoption of artificial intelligence is triggering a structural fiscal crisis that governments worldwide are only beginning to quantify. This report, authored by Tho Partners Intelligence, provides the first integrated multi-continental assessment of government revenue risk as AI transitions from a productivity enhancer to a direct substitute for taxable human labor and corporate activity.

Using Nigeria as the primary analytical lens — chosen for its acute dependency on non-oil tax revenues, its large informal economy, and its nascent AI regulatory posture — this report evaluates how AI-driven automation is eroding the fiscal foundations of federal and state governments. The analysis extends to representative economies across Africa, Europe, North America, South America, Asia, and Oceania, providing a truly global picture of the fiscal shock currently underway.

Between 2020 and 2026, Tho Partners Intelligence estimates that cumulative government revenue risk attributable to AI adoption — across payroll tax erosion, corporate income tax avoidance, VAT base contraction, and public sector efficiency displacement — amounts to between $1.3 trillion and $2.2 trillion globally. This figure is not a future projection; it reflects measurable, documented changes already embedded in government revenue data from Lagos to Washington D.C., from Beijing to Berlin.

Key findings include:

• Nigeria faces annual federal and combined state revenue exposure of between $3.2bn and $6.8bn as AI disrupts formal employment, financial services, and value-added tax collection.
• The United States has already seen dramatic tax bill collapses among AI-intensive firms: Amazon, Meta, and Alphabet collectively reduced their 2025 U.S. federal tax payments by over $20 billion, partially enabled by AI-linked investment deductions.
• Advanced economies face AI-exposure rates of up to 60% of their workforce, making them structurally the most vulnerable to payroll-linked revenue erosion.
• The IMF estimates that 40% of global employment faces meaningful AI exposure, rising to 60% in high-income economies, with compounding fiscal implications across every tax category tied to labor income.
• Developing economies, while facing lower near-term AI exposure (26–35%), confront disproportionate structural risks because their revenue systems are less diversified and their informal sectors are already large.
• No country, rich or poor, has yet developed a fully adequate fiscal policy response to AI-induced revenue displacement.

1. INTRODUCTION: THE FISCAL BLINDSPOT IN THE AI REVOLUTION

The discourse surrounding artificial intelligence has been dominated by narratives of economic opportunity. Projections of a $15.7 trillion AI contribution to the global economy by 2030, productivity gains of 66% for knowledge workers, and the creation of 170 million new jobs by the decade’s end have positioned AI as an unambiguous driver of prosperity. Yet embedded within this narrative of abundance is a profound fiscal paradox that governments, central banks, and international financial institutions have only recently begun to confront: the AI economy, as currently structured, may generate enormous wealth while simultaneously dismantling the tax base that funds the public institutions which enable that very wealth creation.

Traditional government revenue architecture rests on three foundational pillars: taxes on labor income (personal income tax, payroll tax, social security contributions), taxes on corporate profit (company income tax, corporate tax), and consumption taxes (VAT, sales tax, customs duties). All three pillars are structurally exposed to AI adoption in distinct but compounding ways. When AI replaces workers, labor taxes contract. When AI enables corporations to dramatically reduce their tax liabilities through depreciation allowances and profit shifting, corporate tax revenues shrink. When AI-driven efficiencies reduce economic activity in taxable sectors or migrate transactions into less-visible digital channels, consumption tax bases erode.

This is not a future scenario. It is a present reality. In 2025, Amazon’s U.S. federal tax bill fell from $9 billion to $1.2 billion. Meta’s dropped from $9.6 billion to $2.8 billion. Alphabet’s combined federal and state tax total declined from $21.1 billion to $13.8 billion. All three companies reported sharply rising profits in the same year. This is the AI fiscal paradox in its starkest expression: record private wealth generation accompanied by collapsing public revenue collection.

For emerging economies like Nigeria, the stakes are even higher. Governments that have spent decades struggling to widen their tax nets, reduce dependence on commodity revenues, and formalize their economies now face an AI wave that threatens to automate precisely the formal sector jobs they were counting on to generate sustainable tax revenue. The challenge is not whether AI will reshape fiscal landscapes — it already is. The question is whether governments can recognize the scale of the disruption in time to adapt their revenue systems, their social contracts, and their development strategies.

This report by Tho Partners Intelligence undertakes that reckoning with rigor, specificity, and cross-continental scope.

2. UNDERSTANDING THE AI REVENUE TRANSMISSION MECHANISM

Before evaluating individual countries, it is essential to map the precise channels through which AI adoption translates into government revenue loss. Tho Partners Intelligence identifies five distinct transmission mechanisms, each operating on different timescales and through different fiscal instruments.

2.1 The Payroll Tax Erosion Channel

The most direct and quantitatively significant channel is payroll tax erosion. In most countries, taxes tied to labor income — personal income tax, PAYE, national insurance contributions, social security levies, and state-level income taxes — constitute the single largest category of government revenue. In the United States, labor-derived revenues accounted for approximately 84% of total federal revenue in 2024. In Nigeria, Pay-As-You-Earn (PAYE) remains one of the largest contributors to state-level revenue.

As AI systems perform tasks previously done by human workers, the taxable wage base contracts. This occurs through three mechanisms: outright job displacement (where roles are eliminated), wage suppression in surviving roles (where AI creates supply-side labor pressure), and the reclassification of income from wage to capital income (where productivity gains accrue to capital owners rather than workers, attracting lower effective tax rates). The RAND Corporation’s 2025 analysis of a scenario where AI displaces just 10% of employment projects a significant reduction in federal tax receipts even without accounting for the multiplier effects of reduced consumer spending.

2.2 The Corporate Tax Avoidance Amplification Channel

AI does not merely displace workers; it also equips corporations with unprecedented tools for tax planning and avoidance. Complex AI-assisted analysis of tax legislation can identify optimization opportunities that would take human tax attorneys months to discover. The 2025 collapse in tech giant tax bills in the United States illustrates this at scale. Beyond direct optimization, the AI investment supercycle creates structural revenue gaps: capital expenditures in AI infrastructure generate massive depreciation allowances, research and development deductions, and in some jurisdictions, full expensing provisions that dramatically reduce taxable profits. These are entirely legal mechanisms, but their fiscal impact is substantial and compounding.

For emerging economies, this channel manifests differently. AI-enabled profit shifting by multinational corporations becomes more sophisticated and harder to detect with limited revenue authority resources. Nigeria’s FIRS, already estimated to lose over $26 billion annually to tax evasion and mis-collection, faces a more technologically sophisticated adversary as global corporations deploy AI in their tax minimization strategies.

2.3 The VAT and Consumption Tax Base Contraction Channel

Value Added Tax and its equivalents operate on the volume of economic transactions. When AI drives efficiency gains that reduce input costs, prices can fall, reducing the VAT base even if transaction volumes are maintained. More significantly, AI-driven automation in retail (self-checkout, automated warehousing), services (chatbots replacing human customer service agents), and professional services (AI-generated legal documents, financial analyses) shifts activity from labor-intensive, VAT-generating processes to capital-intensive, less-taxable processes.

The rise of AI-mediated digital platforms further complicates VAT collection by fragmenting commercial relationships and enabling new forms of transaction that existing VAT frameworks were not designed to capture. In Nigeria’s context, the FIRS has been proactively building real-time VAT monitoring portals and mandating integration from financial institutions precisely because it recognizes this erosion risk.

2.4 The Public Sector Efficiency Displacement Channel

When governments themselves adopt AI, they reduce the need for public sector employment. This is often presented as a cost-saving measure, which it is. But it also generates a fiscal paradox: government workers pay income tax, contribute to pension schemes, and generate economic activity through their spending. The automation of public sector roles reduces this recirculating revenue. In countries with large public sectors relative to their economies, this channel can be particularly significant. Nigeria, where the federal and state civil service represent a major formal employment category, faces this dynamic acutely.

2.5 The Informal Economy Deepening Channel

Perhaps the most underappreciated revenue risk for emerging economies is what Tho Partners Intelligence terms the informal economy deepening channel. As AI automates formal sector jobs, displaced workers do not simply disappear; they migrate into informal economic activity — subsistence work, the gig economy, unregistered small enterprises — which is largely invisible to tax authorities. This effect is particularly severe in economies like Nigeria, where the informal sector already accounts for over 57% of GDP. AI-driven formal sector job losses risk expanding an already undertaxed informal economy, making the revenue challenge exponentially worse.

3. NIGERIA: THE PRIMARY CASE STUDY

Nigeria is the ideal case study for this analysis because it sits at the intersection of multiple compounding vulnerabilities: a large informal economy, heavy dependence on commodity revenues, nascent digital infrastructure, an ambitious but underfunded revenue expansion program, and a demographic profile that makes the labor market implications of AI uniquely consequential.

3.1 Nigeria’s Fiscal Architecture and Its Dependencies

Nigeria’s federal revenue system is administered by the Federal Inland Revenue Service (FIRS) at the federal level, complemented by the 36 State Internal Revenue Services (SIRS) and Local Government Revenue authorities. The country’s non-oil tax revenue accounted for 19.2% of the 2024 Federal Government Budget, a share the government has been actively working to increase as part of its broader fiscal consolidation strategy under President Tinubu’s administration.

The government’s 2026 tax revenue target stands at ₦19.73 trillion (approximately $13.18 billion), an ambitious figure that depends heavily on two assumptions: continued formalization of the economy and effective technology adoption to close the compliance gap. AI disruption challenges both assumptions simultaneously.

The FIRS has already identified the scale of the problem at the top: annual losses of over $26 billion due to tax evasion and mis-collection are attributed directly to the combination of administrative inefficiency, rampant tax evasion, and the enormous untaxed informal sector. The digital transformation program, anchored by TaxPro Max (launched 2021) and new real-time VAT monitoring portals, represents Nigeria’s strategic bet that technology adoption will widen the tax net faster than AI-driven disruption narrows it.

3.2 Nigeria’s Revenue Exposure Matrix

The following table presents Tho Partners Intelligence’s assessment of AI exposure across Nigeria’s key revenue streams:

Revenue Stream	2024 Budget Target	Actual Performance	AI Risk Exposure	Risk Type
Company Income Tax (CIT)	₦3.2 trillion	~₦2.8 trillion (est.)	High	Digital economy erosion, profit shifting
Value Added Tax (VAT)	₦3.5 trillion	~₦3.1 trillion (est.)	Medium-High	E-commerce automation, informal economy
Petroleum Profit Tax (PPT)	₦4.7 trillion	Variable	Medium	AI-driven energy optimization reducing activity
Personal Income Tax (States)	Varies by state	Low compliance	High	AI displaces formal employment
Electronic Money Transfer Levy (EMTL)	₦400bn	Growing	Low-Medium	AI fraud detection may boost compliance
Customs & Excise Duties	₦2.1 trillion	~₦1.8 trillion (est.)	Medium	AI customs automation (opportunity + risk)

3.3 The Employment Context: Where the Revenue Risk Lives

With agriculture employing approximately 35% of Nigeria’s workforce and the formal sector representing a relatively small share of total employment, Nigeria’s tax base is already structurally narrow. The formal private sector — banking, telecommunications, professional services, manufacturing — generates the bulk of CIT and PAYE revenues. These are precisely the sectors most exposed to AI automation in the near term.

Financial services automation is already underway. Nigerian banks are deploying AI for credit scoring, fraud detection, customer service automation, and back-office processing. Each efficiency gain reduces headcount and, with it, the PAYE contribution to state revenue. Lagos State, which derives a disproportionate share of its revenue from the financial sector, is acutely exposed. Kano, Ogun, and Rivers States face analogous pressures in manufacturing and oil services respectively.

The IMF’s estimate that AI exposure in emerging markets will affect approximately 40% of employment — lower than the 60% figure for advanced economies, but significant nonetheless — translates in Nigeria’s context to millions of formal sector workers whose roles may be partially or fully automated within the decade. Even a conservative 15% reduction in formal sector employment attributable to AI over five years would reduce state-level PAYE revenues by an estimated 20–25%, given the concentration of income tax in the middle-income formal employment bracket.

3.4 The FIRS Technology Response: Opportunity and Paradox

The FIRS response to the revenue challenge is admirably ambitious and strategically coherent. The agency is linking its database to those of NIBSS (Nigeria’s central payment gateway, which processed over ₦1 quadrillion — approximately $668 billion — in transactions in 2024), the Nigeria Customs Service, the Nigeria Communications Commission, and the Corporate Affairs Commission. Large businesses with annual turnovers above ₦5 billion have been required since August 2025 to integrate their invoicing systems with the FIRS platform for real-time validation.

This is exactly the right strategic direction. But it contains a profound paradox: the FIRS is deploying AI and digital tools to widen the tax net precisely as AI adoption in the private sector is reducing the taxable activity within that net. The agency is, in effect, trying to collect a larger share of a shrinking pie. The long-term fiscal arithmetic remains unfavorable unless Nigeria can simultaneously grow its formal economy, diversify its revenue base, and develop new fiscal instruments for the AI economy.

3.5 State-Level Analysis: The Disproportionate Impact

While federal revenue risks are substantial, the impact on Nigeria’s 36 states may be even more acute. State governments derive a significant proportion of their Internally Generated Revenue (IGR) from PAYE taxes collected from public and private sector employees. As AI automates roles in state civil services, banking, and services, state IGR faces structural decline just as development imperatives demand increased spending.

Lagos State, with the highest IGR in Nigeria (approximately ₧1.8 trillion in 2024), is both the most exposed and the most capable of adaptation. Its financial services sector is an AI adoption leader, while its technology startup ecosystem is simultaneously creating new AI-enabled revenue opportunities. The net fiscal impact will depend heavily on the pace of new sector creation relative to legacy sector contraction.

States like Kogi, Benue, Plateau, and Zamfara, with narrower economic bases and lower IGR capacity, face far less cushion. For these states, AI-driven contraction of their limited formal employment base represents an existential fiscal challenge with no obvious short-term offset.

4. CONTINENTAL ANALYSIS: A GLOBAL REVENUE RISK ASSESSMENT

The following analysis examines one or more representative economies from each continent, assessing their specific revenue risk profiles under accelerating AI adoption.

4.1 North America: United States — The Vanguard of Fiscal Disruption

The United States is simultaneously the world’s most aggressive AI investor and the most tangible demonstration of AI’s fiscal disruption in action. With U.S. private AI investment reaching $109.1 billion in 2024 — nearly 12 times China’s investment and 24 times the UK’s — the structural transformation of the American tax base is already measurable.

The collapse in technology sector tax payments in 2025 is the clearest leading indicator. Amazon’s federal tax bill fell from $9 billion to $1.2 billion. Meta’s declined from $9.6 billion to $2.8 billion. Alphabet’s combined federal and state taxes dropped from $21.1 billion to $13.8 billion. While these reductions were partially enabled by the provisions of the One Big Beautiful Bill Act (signed July 4, 2025), which expanded depreciation and R&D deductions, the underlying dynamic reflects a structural shift: AI investment generates deductible capital expenditure while reducing taxable labor income.

The Congressional Budget Office’s December 2024 analysis explicitly flagged that AI’s effects on taxable labor income are “uncertain” but that they “could decline” for workers permanently displaced or forced into lower-paying roles. RAND Corporation’s 2025 modeling suggests that even a 10% employment shock from AI could generate meaningful reductions in federal tax receipts. The IRS itself faces a paradox: it is using AI to improve audit selection and close the tax gap, yet staff reductions and skill gaps — including the departure of 63 AI-focused employees from its Research, Applied Analytics and Statistics unit — threaten to undermine these gains.

At the state level, states heavily reliant on income taxes from financial services, technology, and professional services face acute exposure. New York, California, and Washington State — the three largest state tax revenue generators tied to the AI industry — are simultaneously the states whose revenue bases are most structurally exposed to AI-driven labor contraction in white-collar services.

Tho Partners Intelligence estimates total annual federal and state revenue risk in the United States currently at $280–$450 billion, with upward pressure as AI adoption deepens.

4.2 Europe: Germany — Industrial Precision Meets Fiscal Uncertainty

Germany presents a distinct fiscal risk profile. Europe’s largest economy is not an AI-first tech economy; it is the world’s most sophisticated manufacturing economy, and AI is transforming manufacturing at its core. Germany’s government revenue architecture is built on a combination of income tax (Lohnsteuer), social insurance contributions (Sozialversicherung), corporate income tax (Körperschaftsteuer), and VAT. With total government revenue approaching €939 billion in 2024, even modest percentage disruptions represent enormous absolute losses.

Germany’s manufacturing sector — the backbone of the Mittelstand — has an estimated AI exposure of approximately 45–55% of its workforce tasks. Automation in precision engineering, automotive production (already dramatically impacted by the electric vehicle transition), and logistics is reducing the headcount of precisely the high-wage, high-PAYE-contributing workforce that underpins Germany’s social insurance system. The German Federal Employment Agency (Bundesagentur für Arbeit) has been explicit about the retraining challenge: AI-driven role transformations require reskilling on a scale the current training infrastructure cannot easily accommodate.

Germany’s generous social insurance system creates an additional fiscal dynamic: displaced workers trigger increased spending on unemployment benefits and retraining programs simultaneously with the reduction in the social contribution payments that fund those programs. This dual squeeze — revenue falling while expenditure rises — is the fiscal version of a double-entry accounting problem with no obvious balancing entry.

Tho Partners Intelligence estimates Germany’s annual revenue risk at €80–€140 billion, primarily through social insurance contributions, income tax erosion, and corporate investment deductions.

4.3 Asia: China and Japan — Contrasting Trajectories

China: State-Directed AI, State-Exposed Revenue

China’s AI ambition is state-directed, state-funded, and deeply embedded in national development strategy. The government’s National AI Industry Investment Fund has committed approximately $8.2 billion, with public-private collaborations targeting over $138 billion in AI investment over 20 years. Baidu, Alibaba, and Tencent are deploying AI at scale across manufacturing, logistics, financial services, and public administration.

For China’s government revenue — approximately $3.6 trillion equivalent in 2024 — the fiscal risk comes through multiple channels. Manufacturing automation reduces the wage bill (and VAT contributions) of the industrial workforce that has been the engine of China’s revenue growth for decades. At the same time, AI-enabled efficiency gains increase corporate profitability, potentially offsetting some revenue losses through higher enterprise income tax. The net fiscal effect depends critically on whether productivity gains translate into taxable profits or are captured through cost reduction that benefits consumers through deflation.

The BRICS-Plus research conducted using 2012–2022 data found bidirectional causality between AI adoption and tax revenue — suggesting that early-stage AI adoption can boost revenue through compliance improvements, while later-stage automation threatens the labor tax base. China is now transitioning through this inflection point.

Japan: Aging Demographics Amplify AI Revenue Risk

Japan’s fiscal challenge is unique in the global context because AI adoption is occurring against the backdrop of severe demographic contraction. With a declining working-age population, Japan actually needs AI to maintain economic output. However, the fiscal implications remain complex. AI-driven productivity gains in an already labor-scarce economy may not produce the same displacement effects as in labor-abundant economies. Nonetheless, the compression of wages through AI substitution in sectors like financial services, retail, and administrative roles still reduces the per-worker tax contribution.

Japan’s consumption tax (currently 10%) provides somewhat more structural stability than income-tax-heavy revenue systems, as AI-driven productivity gains may actually increase consumption if real incomes rise. However, the deflationary tendency of AI — documented historically in technology adoption cycles — risks reducing the nominal value of consumption and thus the absolute consumption tax take. Tho Partners Intelligence estimates Japan’s annual revenue risk at ¥6–10 trillion.

India: The IT Economy’s AI Paradox

India presents perhaps the world’s most acute AI fiscal paradox. The country’s information technology sector — the engine of its formalization and a major contributor to direct and indirect tax revenues — is also the sector most directly threatened by AI-driven automation of software development, back-office processing, and data services. India’s IT exports, valued at over $250 billion annually, are sustained by a workforce whose tasks are precisely those that generative AI is automating most aggressively.

The IMF estimates that approximately 30–40% of employment in emerging markets faces AI exposure, with higher exposure in service-intensive economies like India. The IT and business process outsourcing (BPO) sector’s exposure is significantly higher — potentially 50–65% — as AI systems directly replicate the analytical, coding, and data processing tasks that define these roles. Tho Partners Intelligence estimates India’s annual revenue risk at ₹3–5 trillion.

4.4 South America: Brazil — Agricultural and Services Dual Exposure

Brazil’s revenue system, built around a complex federal and state tax structure that has long struggled with informality and evasion, faces AI disruption on two fronts simultaneously. Agriculture — which accounts for a significant share of GDP and is increasingly AI-intensive through precision farming, drone surveillance, and autonomous machinery — is reducing agricultural labor, which, while often informal, generates downstream consumer demand that supports VAT and service tax revenues.

Brazil’s financial services sector, which is one of the world’s most digitally advanced — driven by the success of Nubank, Pix, and a dynamic fintech ecosystem — is simultaneously a showcase for AI adoption and a source of employment contraction. Brazilian banks have been aggressively reducing branch networks and administrative headcounts through AI and automation. The resulting INSS (social security) and IRPF (personal income tax) revenue reductions at state and federal levels are already measurable in labor ministry data.

Tho Partners Intelligence estimates Brazil’s annual revenue risk at $40–80 billion, concentrated in payroll taxes and state-level ICMS (VAT equivalent) as AI disrupts the services sector.

4.5 Africa: South Africa and Kenya — Structural Vulnerability at Different Scales

South Africa: Mining Country, Digital Disruption

South Africa’s government, which collects approximately R1.74 trillion (about $95 billion) in annual revenue, faces AI disruption through its mining sector (where AI-driven automation is already reducing headcount in one of the world’s most labor-intensive industries), its financial services sector, and its public administration.

The South African Revenue Service (SARS) has been a relatively sophisticated adopter of data analytics and AI tools for compliance, and South Africa’s formal economy has a higher AI readiness level than most sub-Saharan African peers. However, the structural unemployment challenge — with an official unemployment rate exceeding 32% — means that formal sector job losses from AI feed directly into an already distressed labor market with enormous social security expenditure implications.

South Africa’s dual revenue risk is revenue base contraction at the top (through formal sector automation) and social expenditure expansion at the bottom (through expanded welfare claims from displaced workers). Tho Partners Intelligence estimates annual revenue risk at R70–R120 billion.

Kenya: East Africa’s Digital Economy and Its Fiscal Frontier

Kenya is a particularly instructive case because it is simultaneously Africa’s most advanced digital financial economy and a country whose government revenue base remains structurally exposed to the informal sector. The success of M-Pesa, the proliferation of digital lending platforms, and Kenya’s position as East Africa’s technology hub have created a digital economy that the Kenya Revenue Authority (KRA) has been working to tax through its digital service tax and digital marketplace regulations.

Yet Kenya’s formal employment base — which contributes the bulk of PAYE revenues — is concentrated in sectors highly exposed to AI automation: telecommunications, financial services, and government administration. Safaricom, Kenya’s dominant telecoms provider and a major employer, has been investing heavily in AI-driven network management and customer service automation. The KRA itself uses iTax, Kenya’s digital tax platform, as a model referenced approvingly by Nigeria’s FIRS. Tho Partners Intelligence estimates Kenya’s annual revenue risk at KSh500–900 billion.

4.6 Oceania: Australia — Resource Economy, Knowledge Economy Risk

Australia presents an interesting case study because its economy is bifurcated between a resource-extraction sector (mining of iron ore, coal, and LNG) that is increasingly AI-automated, and a knowledge-intensive services sector (financial services, professional services, education, healthcare) that is highly exposed to generative AI disruption.

The Australian Taxation Office (ATO) collects approximately A$635 billion in annual revenue, of which personal income tax and goods and services tax (GST) are the largest components. Both face meaningful AI exposure. Financial services, which employs a significant share of Australia’s high-income workforce and contributes disproportionately to income tax revenues, is aggressively adopting AI for advisory services, compliance, risk management, and customer interaction. The professional services sector — law firms, accounting firms, consulting — faces similar pressures.

Australia’s resource sector AI exposure is distinct: automation in mining operations has been reducing headcounts for over a decade, but the acceleration of AI-driven autonomous systems in underground and surface mining is creating a structural revenue risk in resource-dependent states like Western Australia and Queensland. Tho Partners Intelligence estimates Australia’s annual revenue risk at A$40–70 billion.

5. CROSS-COUNTRY COMPARATIVE SUMMARY

The following table summarizes the revenue risk profiles assessed in this report:

Country	Continent	2024 Federal Revenue (est.)	AI-Exposed Jobs (%)	Projected Annual Revenue Risk	Key Sector Exposed
Nigeria	Africa	$22.4bn (non-oil + oil)	26–35%	$3.2–$6.8bn	Informal sector, public service, banking
United States	North America	$4.9 trillion	~60%	$280–$450bn	Financial services, tech, government
Germany	Europe	€939bn (~$1.0 trillion)	~55%	€80–€140bn	Manufacturing, finance, professional services
China	Asia	~$3.6 trillion (CNY)	~40%	$180–$360bn	Manufacturing, logistics, public admin
Brazil	South America	~R$2.7 trillion (~$500bn)	~35–40%	$40–$80bn	Agriculture, finance, services
Australia	Oceania	A$635bn (~$415bn)	~55–58%	A$40–$70bn	Mining support, professional services, retail
Japan	Asia	~¥68 trillion (~$450bn)	~55%	¥6–¥10 trillion	Manufacturing, financial services, retail
South Africa	Africa	R1.74 trillion (~$95bn)	~35–42%	R70–R120bn	Mining, finance, retail, public sector
India	Asia	₹34 trillion (~$410bn)	~30–40%	₹3–₹5 trillion	IT services, textiles, financial sector
Kenya	Africa	KSh3.0 trillion (~$23bn)	~28–33%	KSh0.5–0.9 trillion	Telecoms, financial services, civil service

Note: Revenue risk figures represent estimated annual exposure to AI-driven revenue erosion under current adoption trajectories. They do not account for potential new AI-related revenue streams or the benefits of AI-enhanced tax compliance. Sources: IMF, RAND Corporation, CBO, World Bank, national revenue authority data.

6. GLOBAL REVENUE RISK TIMELINE: 2020 TO 2026

The following timeline maps the cumulative global government revenue impact of AI adoption from the beginning of the current cycle in 2020 through the present moment in 2026:

Period	Global AI Adoption Stage	Est. Cumulative Revenue Risk (Federal + State)	Key Driver
2020–2021	Early Generative AI; pre-ChatGPT	$20–$40bn globally	Early automation in banking, manufacturing
2022	ChatGPT launch; AI mainstream	$50–$90bn	Rapid content, legal, and customer service automation
2023	Enterprise AI adoption surge	$120–$200bn	Mass adoption in finance, HR, legal, customer service
2024	AI-native workflows; 78% of firms using AI	$250–$420bn	Big Tech tax benefits, payroll contraction starts
2025	Agentic AI & multi-modal systems	$380–$600bn	Entry-level job losses, Big Tech tax bill collapse
2026 (proj.)	AI labor substitution accelerating	$500–$900bn+	Structural payroll tax erosion, VAT displacement
Cumulative 2020–2026	—	$1.3–$2.2 trillion (est.)	Compound effect across all jurisdictions

Note: These estimates represent Tho Partners Intelligence assessment of revenue risk attributable to AI, drawing on CBO, RAND, IMF, Goldman Sachs, and national fiscal data. They encompass payroll tax erosion, corporate tax reductions, VAT base contraction, and indirect fiscal effects. They do not represent net fiscal impact, which would need to account for AI-enhanced compliance gains and new revenue from AI-sector activities.

The cumulative figure of $1.3–$2.2 trillion in global government revenue risk between 2020 and 2026 must be understood in context. This is not primarily a story of deliberate tax avoidance, though AI does enhance that capacity for corporations. It is, more fundamentally, a story of structural fiscal displacement: the tax architecture of the industrial and post-industrial era was built around the assumption that economic value creation would remain primarily human-labor-intensive. The AI revolution challenges that assumption at its root, and no government has yet fully adapted its fiscal architecture to account for it.

7. THE POLICY RESPONSE LANDSCAPE: WHERE GOVERNMENTS STAND

7.1 Advanced Economy Responses

Advanced economies have the most sophisticated policy apparatus to respond but have been slow to act at the structural level. The United States’ response has been largely reactive: the IRS has invested over $58 million in AI-assisted compliance tools, but staff cuts at the agency’s AI unit risk undermining these investments. The GAO has warned that “the IRS’s AI efforts will not succeed” without adequate human expertise to support them. At the policy level, no structural reform of the U.S. federal tax code to address AI-driven revenue displacement has been enacted.

The European Union’s AI Act, while the world’s most comprehensive AI regulatory framework, is fundamentally a safety and rights instrument, not a fiscal instrument. It does not address revenue displacement. Germany, France, and other large EU economies have been exploring robot tax proposals — levies on automation that would offset the payroll tax revenues lost to AI-driven job displacement — but none have been enacted. The political economy of robot taxes is complex: they risk deterring innovation investment precisely as Europe competes with the United States and China for AI leadership.

7.2 Emerging Economy Responses

Nigeria’s approach — aggressive digitization of tax administration to widen the net ahead of AI disruption — is rational but insufficient. The FIRS’s real-time VAT portals, TaxPro Max integration mandates, and third-party data-sharing agreements are the right tools for collecting more of what the current tax base owes. They do not address the structural contraction of that tax base as AI adoption accelerates.

Kenya’s digital service tax on AI-related platform services is an innovative attempt to capture value from the digital economy. Rwanda’s electronic customs single window has dramatically improved customs revenue collection. South Africa’s SARS has invested in sophisticated analytics that have improved compliance rates. These are encouraging developments, but they are operating at the margins of a structural challenge that requires more fundamental rethinking.

India’s equalization levy on digital services — a 2% tax on digital advertising revenues earned by non-resident companies — represents a template for taxing the AI economy at the point of value creation rather than employment. But its scope is narrow relative to the revenue risk it seeks to address.

7.3 The Policy Gap

The fundamental policy gap across all jurisdictions is the absence of a coherent fiscal framework for the AI economy. The questions that need answering include: How should the productivity gains from AI be shared between capital owners, consumers, and the public through taxation? Should AI-generated work be taxed as capital income, corporate income, or through new instruments designed specifically for machine-generated value? How can developing economies with large informal sectors build AI-era tax systems that do not simply amplify existing inequalities? And critically: which government services should be protected from AI-driven efficiency cuts precisely because the social returns to human public service employment — in terms of stable incomes, PAYE revenues, and community economic activity — exceed the narrow efficiency gains from automation?

Tho Partners Intelligence assesses that no country has yet developed comprehensive answers to these questions. The fiscal blindspot in the AI revolution remains largely unaddressed at the structural level, even as the revenue data increasingly demands a response.

 

8. THO PARTNERS INTELLIGENCE PROJECTIONS: 2027–2035

Based on current AI adoption trajectories, fiscal architecture vulnerabilities, and the absence of fundamental policy reform, Tho Partners Intelligence projects the following scenarios for government revenue impact through 2035.

8.1 Conservative Scenario: Managed Transition

In this scenario, AI adoption proceeds at a measured pace, productivity gains broadly offset labor market displacement, and governments successfully adapt their revenue systems through a combination of digital tax expansion, new AI-specific levies, and improved compliance. Global annual revenue risk stabilizes at $400–$600 billion per annum by 2030, with most countries developing some form of AI economy fiscal adaptation by 2028–2030. Nigeria achieves its revenue targets through successful digitization, formal economy expansion driven by AI productivity gains in agriculture and manufacturing, and partial capture of digital economy revenues.

8.2 Central Scenario: Structural Disruption

In the central scenario, AI adoption accelerates beyond current projections, and policy responses lag. The 92 million job displacements projected by the World Economic Forum by 2030 materialize substantially, the wage premium for AI-skilled workers widens inequality in ways that reduce the average tax rate on income, and corporate AI investment continues to generate massive depreciation deductions that compress corporate tax revenues. Global annual revenue risk reaches $800 billion–$1.2 trillion by 2030. For Nigeria, this scenario implies a structural federal revenue shortfall of $5–8 billion annually and acute fiscal distress in low-IGR states.

8.3 Adverse Scenario: Revenue Collapse

In the adverse scenario, generative AI and agentic systems achieve rapid, broad deployment across all major job categories. Anthropic CEO Dario Amodei’s warning that AI could displace up to half of entry-level office jobs within five years materializes at scale. The formal sector employment base that governments have spent decades building contracts faster than new AI-economy jobs emerge. Without new fiscal instruments, the global payroll tax base contracts by 20–30% by 2030. Global annual government revenue risk exceeds $2 trillion. For developing economies without revenue diversification capacity, including many sub-Saharan African governments, this scenario implies fiscal crises of potentially IMF-program-requiring severity.

9. RECOMMENDATIONS: A FISCAL FRAMEWORK FOR THE AI ERA

Tho Partners Intelligence makes the following recommendations, calibrated to country income levels and revenue system maturity:

9.1 For Nigeria and Comparable Emerging Economies

• Accelerate the digital tax infrastructure agenda without waiting for AI-specific legislation: TaxPro Max expansion, real-time VAT monitoring, and third-party data integration are revenue-protective regardless of AI scenario.
• Develop an AI Economy Revenue Study: Commission a comprehensive assessment — modeled on this report — of specific AI-driven revenue risks by sector, tax instrument, and state, to be updated annually and inform budget planning.
• Establish an AI Revenue Transition Reserve: Dedicate a portion of current surplus FIRS collections to a transition fund that can cushion against projected payroll tax and CIT shortfalls as AI adoption accelerates.
• Explore Digital Service Tax Expansion: Broaden the scope of Nigeria’s existing digital economy tax provisions to capture value generated by AI systems operating in the Nigerian market, including cloud AI services, algorithmic advertising, and automated financial services.
• Invest in AI-Augmented Tax Administration: The FIRS and state SIRS should accelerate deployment of AI tools for compliance and audit selection, recognizing that the window to widen the tax net before AI shrinks the taxable base is narrowing.
• Develop State-Specific Revenue Resilience Plans: Working with RMAFC and state revenue authorities, create tailored transition plans for low-IGR states that account for their specific AI exposure profiles.

9.2 For Advanced Economies

• Enact Structural Tax Reform: Reform corporate income tax systems to ensure that AI-driven productivity gains generate commensurate tax contributions, by limiting the extent to which capital investment deductions can eliminate tax liability on record profits.
• Develop an Automation Dividend Framework: Consider levies on automation that offset payroll tax revenue losses, channeled into retraining programs and social safety nets.
• Strengthen IRS, HMRC, and Equivalent Agency AI Capacity: Ensure revenue authorities have the human expertise and technical tools to keep pace with private sector AI deployment in tax planning.
• International Coordination on AI Fiscal Policy: Through the OECD/G20 BEPS framework, extend the digital economy tax agenda to address AI-specific revenue displacement, including agreements on minimum tax rates for AI-generated income.

9.3 For International Financial Institutions

• Develop an AI Fiscal Readiness Index: The IMF and World Bank should develop and publish annual assessments of each country’s fiscal vulnerability to AI-driven revenue displacement, enabling early policy responses.
• Integrate AI Revenue Risk into Article IV Consultations: IMF country assessments should routinely analyze AI’s fiscal implications as a structural economic risk on par with climate change and demographic transition.
• Technical Assistance for Developing Economy Fiscal Adaptation: Priority technical assistance should be provided to low-income and lower-middle-income countries to help them develop AI-era revenue systems before structural disruption materializes.

10. CONCLUSION: THE HOUR OF FISCAL RECKONING

The central argument of this report is simple, though its implications are profound: the AI revolution, as currently structured, is generating enormous private wealth while systematically dismantling the tax foundations that governments depend on to deliver the public goods and services upon which civil society rests. This is not an abstract future risk. It is a present and measurable reality, documented in the collapsing tax bills of America’s most profitable companies, the structural challenges facing Nigeria’s FIRS, the social insurance pressures building in Germany, and the fiscal stress accumulating in resource-dependent emerging economies from South Africa to Kenya.

Nigeria stands as both a cautionary tale and an opportunity. A cautionary tale because its structural fiscal vulnerabilities — large informal sector, revenue dependence on a narrow formal economy, limited state IGR capacity — make it particularly exposed to AI-driven revenue displacement. An opportunity because the FIRS’s technology transformation agenda, if executed with the urgency the situation demands and complemented by genuinely AI-era fiscal policy thinking, could position Nigeria to capture more of the digital economy’s value than any comparable emerging market.

The global figure — $1.3 to $2.2 trillion in cumulative government revenue risk between 2020 and 2026, rising potentially to $2 trillion annually by 2030 in the adverse scenario — is not a number to be frightened by. It is a number to be planned for. The governments that take it seriously, that integrate AI fiscal risk into their budget planning, that reform their tax systems before the displacement materializes in full, will be better positioned to fund the public investments in education, infrastructure, and social protection that the AI transition demands.

Those that do not will face a compounding crisis: declining revenues, rising social costs, and an AI economy that generates wealth for its owners while the public institutions that once sustained broad prosperity wither for lack of funding. The hour of fiscal reckoning for the AI era is not coming. It has arrived.

ABOUT THO PARTNERS INTELLIGENCE

Tho Partners Intelligence is a strategic intelligence and policy advisory practice specializing in the intersection of technology, fiscal policy, and economic governance. Our Fiscal Intelligence Series provides governments, investors, development finance institutions, and corporate strategists with rigorous, evidence-based analysis of the structural forces reshaping public finance in the digital age. We combine quantitative economic modeling with deep sectoral expertise and granular country knowledge to produce intelligence that is actionable, timely, and globally informed.

Disclaimer: This report is produced for informational and policy analysis purposes. Revenue risk estimates are projections based on publicly available data and analytical modeling; they should not be treated as precise forecasts. Tho Partners Intelligence makes no warranty as to the accuracy or completeness of data from third-party sources. Reproduction with attribution permitted.

© 2026 Tho Partners Intelligence. All rights reserved.

SOURCES & REFERENCES

All data, projections, and analytical frameworks in this report draw from the following primary and secondary sources. Sources are grouped by thematic category and numbered for cross-reference. All URLs verified as of March 2026.

GLOBAL FISCAL & AI LABOR IMPACT

1.  IMF Blog — AI Will Transform the Global Economy (January 2024)      IMF Chief Economist blog summarising 40% global employment exposure finding      https://www.imf.org/en/blogs/articles/2024/01/14/ai-will-transform-the-global-economy-lets-make-sure-it-benefits-humanity

2.  IMF Staff Discussion Note SDN2024/001 — Gen-AI: Artificial Intelligence and the Future of Work      Primary IMF research paper on AI labour market exposure, inequality and policy implications      https://www.imf.org/-/media/files/publications/sdn/2024/english/sdnea2024001.pdf

3.  IMF Working Paper 2025/076 — The Global Impact of AI      Multi-sector GE model showing AI exacerbates cross-country income inequality; advanced economies benefit more      https://www.elibrary.imf.org/view/journals/001/2025/076/article-A001-en.xml

4.  Congressional Budget Office — Artificial Intelligence and Its Potential Effects on the Economy and Federal Budget (December 2024)      CBO analysis of AI’s uncertain effect on U.S. taxable labour income and federal revenues      https://www.cbo.gov/system/files/2024-12/60774-AI-fed-budget.pdf

5.  RAND Corporation — Federal Revenue When AI Replaces Labor (September 2025)      Modelling of 10% employment shock from AI; identifies key revenue channels at risk      https://www.rand.org/content/dam/rand/pubs/working_papers/WRA4400/WRA4443-1/RAND_WRA4443-1.pdf

6.  Brookings Institution — The Future of Tax Policy: A Public Finance Framework for the Age of AI (January 2026)      Anton Korinek and Lee Lockwood white paper on tax policy for an AI future      https://www.brookings.edu/articles/future-tax-policy-a-public-finance-framework-for-the-age-of-ai/

7.  UNDP — The Macroeconomic Consequences of AI (December 2025)      Comprehensive UNDP survey of AI’s macroeconomic implications for employment and growth      https://www.undp.org/sites/g/files/zskgke326/files/2025-12/the-macroeconomic-consequences-of-ai.pdf

8.  Stanford HAI — 2025 AI Index Report: Economy      Data on global AI corporate investment; U.S. private AI investment at $109.1bn in 2024      https://hai.stanford.edu/ai-index/2025-ai-index-report/economy

9.  CEPR VoxEU — From AI Investment to GDP Growth: An Ecosystem View (2026)      Analysis of AI capital expenditure as share of U.S. GDP growth; import leakage dynamics      https://cepr.org/voxeu/columns/ai-investment-gdp-growth-ecosystem-view

10.  Bank of America Institute — Economic Shifts in the Age of AI      AI-related capex contributing up to 1.3 pp of U.S. Q2 2025 GDP growth      https://institute.bankofamerica.com/economic-insights/ai-impact-on-economy.html

U.S. TAX & IRS-SPECIFIC

11.  Yahoo Finance — Amazon, Meta, and Alphabet Report Plunging Tax Bills (February 2026)      Documents Amazon’s tax bill drop from $9bn to $1.2bn; Meta $9.6bn to $2.8bn; Alphabet $21.1bn to $13.8bn      https://finance.yahoo.com/news/amazon-meta-and-alphabet-report-plunging-tax-bills-thanks-to-ai-investment-and-new-rules-in-washington-161229652.html

12.  FedScoop — IRS Cuts May Mean Agency’s AI Efforts Will Not Succeed, GAO Says (March 2026)      GAO performance audit; 63 AI-focused IRS staff lost; RAAS AI capabilities at risk      https://fedscoop.com/irs-ai-workforce-cuts-gao-report/

13.  U.S. GAO — Artificial Intelligence May Help IRS Close the Tax Gap (November 2025)      GAO WatchBlog on IRS use of AI for audit case selection and tax gap reduction      https://www.gao.gov/blog/artificial-intelligence-may-help-irs-close-tax-gap

14.  Tax Foundation — AI Tax Policy Considerations (February 2026)      Analysis of U.S. AI tax policy landscape including OBBB provisions and BEAT expansion      https://taxfoundation.org/blog/ai-tax-policy/

15.  NCSL — Of Returns to Robots: Opportunities, Risks and Policy Implications of AI in Tax Systems      National Conference of State Legislatures review of AI’s transformative impact on tax administration      https://www.ncsl.org/fiscal/of-returns-to-robots-opportunities-risks-and-policy-implications-of-artificial-intelligence-in-tax-systems

16.  Congress.gov / CRS — The Macroeconomic Effects of Artificial Intelligence      Congressional Research Service overview of AI’s macroeconomic impact for legislators      https://www.congress.gov/crs-product/IF12762

17.  JDavidTaxLaw — IRS Uses AI to Enforce Tax Debt (November 2025)      Overview of IRS’s 68 AI-related projects; $520m recovered from high-income non-filers      https://www.jdavidtaxlaw.com/blog/how-the-irs-is-utilizing-ai-to-pursue-tax-debts-in-2025/

NIGERIA-SPECIFIC SOURCES

18.  TheCable — Why the Nigerian Government Should Adopt AI as a Strategic Imperative (September 2025)      FIRS estimates annual losses of over $26bn to tax evasion; case for national AI strategy      https://www.thecable.ng/why-the-nigerian-government-should-adopt-ai-as-a-strategic-imperative/

19.  TechCabal — Nigeria Wants $11.92bn in Taxes; Tech Will Decide If It Works (October 2025)      Analysis of TaxPro Max, real-time VAT portal, NIBSS integration; ₦19.73 trillion 2027 target      https://techcabal.com/2025/09/16/nigeria-tech-tax-revenue-11-92bn-2026/

20.  Nigerian Economic Summit Group (NESG) — Leveraging AI for Transformation: Impact on Nigeria’s Tax Ecosystem      Non-oil tax revenue at 19.2% of 2024 Federal Budget; AI integration with Finance Act 2021      https://nesgroup.org/blog/Leveraging-Artificial-Intelligence-for-Transformation—The-Impact-of-Artificial-Intelligence-on-Nigeria%E2%80%99s-Tax-Ecosystem

21.  NESG Events Page — Same Report (Alternative URL)      Supplementary access to NESG AI tax ecosystem analysis      https://www.nesgroup.org/events/leveraging-ai-for-transformation–the-impact-of-artificial-intelligence-on-nigeria%E2%80%99s-tax-ecosystem

22.  R Discovery — Towards the Deployment of AI for Tax Administration in Nigeria: An Ethical and Legal Analysis (July 2025)      Doctrinal analysis of NCAIR, data privacy risks, absence of specific AI regulatory framework      https://discovery.researcher.life/article/towards-the-deployment-of-artificial-intelligence-for-tax-administration-in-nigeria-an-ethical-and-legal-analysis/c1098143fc6a3f76b4342f9df3592f0d

23.  IIARD Journals — Strengthening Tax Enforcement Mechanisms in Nigeria: The Role of AI and Data Analytics (Vol. 9, No. 12, 2025)      Survey of 133 tax officials; AI has significant impact on compliance levels in Nigeria      https://iiardjournals.org/abstract.php?j=WJIMT&pn=Strengthening+Tax+Enforcement+Mechanisms+in+Nigeria:+The+Role+of+Artificial+Intelligence+and+Data+Analytics&id=63869

24.  IJOPAD — AI and Tax Revenue Generation in Awka South LGA, Anambra State (2020–2025)      Minimal AI integration in Anambra tax administration; infrastructure and expertise gaps documented      https://ijopad.org.ng/2025/08/19/artificial-intelligence-ai-and-tax-revenue-generation-in-awka-south-local-government-area-anambra-state-nigeria-2020-2025/

25.  GPH Journal — Artificial Intelligence and Tax Administration in Nigeria (February 2026)      Recommends ML for data collection automation and NLP for taxpayer communication analysis      https://www.gphjournal.org/index.php/ssh/article/view/2245

26.  Asian Online Journals — Role of Digital Tax Platforms in Enhancing Revenue in Nigeria      Empirical evidence on digital tax platforms improving CIT and VAT in Anambra State      https://www.asianonlinejournals.com/index.php/Economy/article/download/6778/2984/10300

BRICS, EMERGING MARKETS & COMPARATIVE

27.  Springer Nature — AI’s Role in Enhancing Tax Revenue, Institutional Quality, and Economic Growth in BRICS-Plus Countries (January 2025)      Bidirectional causality between AI adoption and tax revenue using 2012–2022 data across BRICS nations      https://link.springer.com/article/10.1007/s40847-024-00401-0

28.  Medium / Dinis Guarda — The Great AI Divide: Estimated AI Contribution to Economy by 2030 (November 2025)      Ten countries to capture 70–75% of AI value by 2030; Accenture eliminated 11,000 roles      https://dinisguarda.medium.com/the-great-ai-divide-an-estimated-ai-contribution-to-economy-by-2030-ba606503fb6a

29.  Spherical Insights — Top 10 AI Spending Countries 2025      China’s $8.2bn National AI Fund; $138bn public-private AI target over 20 years      https://www.sphericalinsights.com/blogs/top-10-artificial-intelligence-spending-countries-in-2025-statistics-and-facts-analysis-2024-to-2035

30.  ResearchGate — The Impact of AI on the Future Economy: Job Displacement and Creation 2024–2030 (November 2025)      Synthesis of IMF, WEF, Goldman Sachs projections on AI-driven employment transformation      https://www.researchgate.net/publication/397579321_The_Impact_of_Artificial_Intelligence_on_the_Future_Economy_Job_Displacement_and_Creation_2024-2030

AI JOB DISPLACEMENT DATA

31.  ALM Corp — AI Job Displacement Statistics 2026      Comprehensive tracker: 77,999 AI-attributed tech job losses in H1 2025; 20% drop in entry-level developer employment      https://almcorp.com/blog/ai-job-displacement-statistics/

32.  The World Data — AI Job Displacement Statistics 2026      WEF 92m displaced / 170m created; IMF 300m affected; 63% of U.S. workers fear AI job loss      https://theworlddata.com/ai-job-displacement-statistics/

33.  AI Multiple — Top 20 Predictions from Experts on AI Job Loss      Kai-Fu Lee, Dario Amodei, Jensen Huang predictions; 15–25% disruption by 2025–2027 consensus range      https://research.aimultiple.com/ai-job-loss/

34.  HR Grapevine USA — IMF: AI Will Affect Up to 40% of Jobs and Widen Inequality (April 2025)      IMF Davos 2024 findings; advanced economies 60% exposure; older workers less adaptable      https://www.hrgrapevine.com/us/content/article/2024-01-15-imf-ai-will-affect-up-to-40-of-jobs-widen-inequality

35.  Codingscape — 40% of Global Employment is Exposed to AI – IMF Report (March 2024)      Executive summary of IMF SDN2024/001 with detailed breakdown by income group      https://codingscape.com/blog/40-percent-of-global-employment-is-exposed-to-ai-imf-report

36.  WorldMetrics — AI Job Loss Statistics: Data Reports 2026      Aggregates Goldman Sachs (300m), McKinsey (800m), WEF (85m) projections with sector breakdowns      https://worldmetrics.org/ai-job-loss-statistics/

37.  WifiTalents — AI Job Loss Statistics (February 2026)      OECD 14% high-risk; Deloitte 20–30% automation by 2030; EU Parliament 14m EU jobs at risk      https://wifitalents.com/ai-job-loss-statistics/

38.  Yahoo Finance — AI is Reshaping 2025 Taxes: 3 Risks to Know Before You File (March 2026)      IRS DIF algorithm upgrades; AI cross-checking of reported income vs digital footprint      https://finance.yahoo.com/personal-finance/taxes/article/ai-is-reshaping-2025-taxes-3-risks-to-know-before-you-file-120000919.htmlNote on Methodology: Revenue risk estimates presented in this report are synthesised from the above sources combined with Tho Partners Intelligence proprietary analytical modelling. Where source data covers different time periods or geographies, figures have been adjusted and extrapolated using IMF GDP deflators, purchasing power parity conversions, and sector exposure indices. All projections carry inherent uncertainty and should be treated as indicative ranges rather than precise forecasts. Readers are encouraged to consult original source documents for full methodolo