evidoria

← Back to browse

Good practice Imported

Mozambique's AI-Driven Tax Modernisation Roadmap — A 2028 Plan to Lift Revenue via Machine-Learning Audits

Mozambique · Maputo · See the Mozambique profile · See the Maputo profile

Evidence: Descriptive / self-reported Top 83% 40/100 · Ask Evidence Copilot about this practice

Mozambique's Finance Ministry approved a fiscal roadmap (CFMP, June 2025) using AI — ML anomaly detection, NLP audits, digital-twin modelling — to lift tax revenue from 417.4bn to 492.2bn meticais by 2028. Still unimplemented: no AI system is deployed, no independent results.

417.4 billion meticais
Projected state tax revenue (2026)
492.2 billion meticais
Projected state tax revenue (2028)
25.4 %
Projected tax revenue as share of GDP (2026)
25.7 %
Projected tax revenue as share of GDP (2028)

Details

Maturity
Pilot
Promoter
Ministério da Economia e Finanças de Moçambique (MEF)
Period
2025-2028
Keywords
tax administration, VAT compliance, fraud detection, fiscal policy, machine learning

Context

Mozambique's Ministerio da Economia e Financas (MEF) approved, around June 2025, the 2025-2027 Cenario Fiscal de Medio Prazo (Medium-Term Fiscal Scenario, CFMP), setting out a roadmap for the country's tax authorities to modernise value-added-tax administration and lift compliance through 2028.

Activities

The roadmap describes a shift from manual, reactive audits toward AI-assisted, risk-based enforcement: machine-learning models to monitor transactions and flag anomalies in real time, natural-language-processing tools to automate document review, generative-AI chatbots to guide taxpayers through filing requirements, and digital-twin simulations to model how proposed tax changes would affect revenue. The plan also covers broader measures -- revising VAT and corporate-tax codes, cutting tax exemptions and formalising informal-sector taxpayers -- alongside these AI components.

Results

The CFMP projects state tax revenue rising from 417.4 billion meticais (about EUR 5.5 billion, 25.4% of GDP) in 2026 to 492.2 billion meticais (about EUR 6.5 billion, 25.7% of GDP) by 2028.

Conclusions

These are fiscal-planning targets tied to the reform package as a whole, not measured results attributable to the AI tools specifically -- as of this writing no AI system described in the roadmap has been deployed, and no independent evaluation of its accuracy, bias or actual revenue effect has been published.

Implementation
Indicative cost
Medium (€50k–€500k) — Fiscal-planning roadmap; specific AI-component budget not itemised.
Time to results
Long (> 3 years) — Approved around June 2025; roadmap targets run through 2028.
Staffing & skills
Ministerio da Economia e Financas (MEF), Mozambican tax authorities

Conditions for success

  • Roadmap bundles AI tools with broader tax-code and enforcement reforms rather than treating AI as a standalone fix

Common failure modes

  • No AI system described in the roadmap has yet been deployed
  • No independent evaluation of accuracy, bias or actual revenue effect has been published

Where it fits

Governance type
national ministry of finance roadmap
Scale
national (Mozambique)
Income level
low-income

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Similar practices you may find useful