Predictive analysis and machine learning tackles €185 million VAT and other tax frauds
The Austrian Federal Ministry of Finance’s specialised Artificial Intelligence tax fraud department, Predictive Analytics Competence Center (PACC), has been enjoying success in identifying various tax frauds. PACC’s work includes analysing VAT missing-trader fraud, especially via SAF-T, and support services in the area of data management and combating fraud.
Many other examples of tax authorities adopting AI VAT fraud measures have been published.
Watch VATCalc’s new AI VAT Advisor deliver tax advice (legislative references and tax cases) on complex VAT questions.
The centre has adopted a range of AI techniques, including:
- Predictive Analytics – using established AI and machine learning models for pre-processing data
- Advanced Analytics – other analysis, including Austrian SAF-T based audits.
- Tax Analytics – looks a consolidating wider tax data, including salary data.
- Customs Analytics – risk profiling; customs-related support; and international fraud cases.
The Austrian Tax Fraud Investigation Unit in May took part in a pan-EU operation to bring to justice a €195 million VAT fraud. Fraud identified included a fake cross-border seller of pallets in carousel fraud (a complex missing trader category of unlawfully collecting VAT on zero-rated intra-community supplies). These fraud schemes use the peculiarity of EU trade, in which cross-border transactions are VAT-free by obtaining an unlawful VAT refund from the state after a first tax-free sale.
The PACC has rolled out various programs to address risks in tax audits and customs control. These initiatives include analyzing past data (ex-post), real-time analysis (ex-ante), and forward-looking assessments to identify risks in advance. With the rise of digitalization, cryptocurrencies, and global data exchange, the PACC is focusing on improving the efficiency of tax administration.
In the realm of real-time audits, machine learning is now being used to assess risks and better select audit cases. This includes scrutinising new businesses to flag potential tax evaders early on helping to identify missing trader VAT fraud and close the VAT Gap. The PACC also leverages data from e-commerce platforms and international tax information exchanges to combat fraud, such as missing-trader fraud.
These projects highlight the PACC’s commitment to maintaining a modern and efficient tax administration, capable of meeting the challenges of the digital economy and global financial markets. The PACC’s work ensures that public funds are used wisely, supporting services like security, education, and infrastructure, while cracking down on tax evaders so that honest taxpayers are not unfairly burdened.
Looking ahead, PACC plans to further enhance its analytical capabilities. This includes incorporating more advanced AI, like generative models AI for VAT and strengthening partnerships both domestically and internationally.
