OECD’s 2026 Anti‑Corruption and Integrity Outlook
The OECD’s annual Anti-Corruption and Integrity Outlook 2026 provides a global assessment of how countries are designing and implementing systems to combat fraud, corruption and integrity risks across public institutions. Covering 37 OECD and 25 partner countries, it draws on new data to evaluate anti-corruption frameworks and highlights emerging risks, including fraud, public procurement exposure and organised crime.
This year’s report emphasises the role of AI in fraud detection by tax authorities, but also, for VAT, e-invoicing and e-reporting.
See how VATCalc is already using AI tax code mapping and AI VAT advice. VATCalc has developed the first VAT-dedicated agentic AI service for calculations and returns.
AI pays major role in tax fraud detection
The OECD’s March 2026 report confirms that artificial intelligence is now a core tool for tax authorities. Across surveyed administrations, 76% use AI for enforcement and 69% for risk assessment, with the primary focus on detecting fraud and evasion through pattern analysis across large datasets.
This sits within a broader shift outlined in the OECD’s report. Fraud is identified as one of the fastest-growing risks, with organisations estimated to lose around 5% of funds annually. Tax authorities are therefore moving towards risk-based, data-led enforcement, supported by digital systems and improved data interoperability.
For VAT, this trend is accelerated by e-invoicing and digital reporting regimes, which provide authorities with transaction-level data in near real time. AI enables the identification of anomalies such as input/output mismatches, supply chain inconsistencies and potential carousel fraud. AI use by VAT authorities has been stepping up, the report notes.
Friction in AI adoption
However, implementation challenges remain. The OECD highlights a 19% gap between policy and execution, driven by limited data access, legal constraints on data sharing, and shortages of technical expertise. In addition, many authorities lack robust frameworks to measure the effectiveness and return on investment of AI deployment.
Transparency is also critical. AI-driven decisions must be explainable and supported by clear audit trails to withstand legal and administrative scrutiny.
The direction is clear. VAT compliance is moving towards continuous, data-driven audit environments. Businesses will need accurate, transaction-level VAT determination and defensible logic aligned to legislation to respond effectively to increasingly targeted enforcement.
