Predictive models rank taxpayers/returns by non-compliance probability and expected audit yield
- Who we serve
- Governments and Public Agencies
- Category
- PFM, Tax & Customs
- Value
- Risk management
- Human role
- Human approves every decision
- Maturity
- B-READY topics
- Taxation
The problem
Without risk-based audit case selection, revenue authorities rely on sampling and after-the-fact review, while the signals in tax returns, invoices, customs declarations, third-party and financial data sit unread, so problems surface only after the loss, incident or complaint.
How it works
Risk-based audit case selection - Predictive models rank taxpayers/returns by non-compliance probability and expected audit yield. It draws on tax returns, invoices, customs declarations, third-party and financial data and is typically owned by revenue authorities. Where the human comes in: revenue authorities review and approve before any action affects a person, a payment or a case; the AI flags, ranks and drafts. It is scaled at leading organisations and piloted elsewhere; examples include UK HMRC Connect; Australia ATO; several African revenue authorities via ATAF.
Where the human comes in
Revenue authorities review and approve before any action affects a person, a payment or a case; the AI flags, ranks and drafts
What it needs
tax returns, invoices, customs declarations, third-party and financial data
What to plan for
Feedback loops bias toward previously audited segments; human review of selections
In the field
UK HMRC Connect; Australia ATO; several African revenue authorities via ATAF
OECD Tax Administration 3.0 (2020); IMF TNM 2024/006 (2024)
Public deployments by other organizations, listed for context. Not Sphere projects.