AI@AUEB: Machine Learning for VAT fraud detection

ΑΙ@AUEB talk, NEW DAYTUESDAY 23/5/2023, 16:15 - 17:30

Room T105, Troias Building

Speaker: Angelos Alexopoulos (Dept. of Economics, AUEB)

Title: A machine learning and network approach to Value Added Tax fraud detection

Value Added Tax (VAT) fraud erodes tax revenues and puts legitimate businesses at a  disadvantaged position thereby impacting inequality. Identifying and combating VAT fraud before it occurs is therefore important for welfare. This paper proposes suitably flexible machine learning algorithms for detection of VAT fraudulent transactions. The innovation of the algorithms is in utilising the information provided by their VAT structure. Making use of the universe of Bulgarian VAT data the fraud detection algorithms detect around 50 percent of the VAT fraud and outperform well-known techniques that ignore the  network of VAT transactions. Importantly, the proposed methods are automated, and can be implemented following the taxpayers' submission of their VAT returns, enabling tax revenue authorities, in the EU and elsewhere, to prevent large losses of tax revenues through performing early identification of fraud between business-to-business transactions within the VAT system. The approach developed can be applied to similar problems where network interactions exist and can be quantified, such as fraud in online transactions, in insurance and social security systems as well as in telecommunication systems.

and virtually via MS Teams here.

To subscribe to the mailing list of AI@AUEB, send a message (with any subject and body) to If you have an AUEB account and want to view all scheduled AI @ AUEB Lecture Series in your MS Teams calendar, subscribe to the "AI@AUEB" group on MS Teams (code: r2dtl45). Team members can also send text messages (chat) to other team members.

If you are an AI researcher or practitioner, please consider becoming a member of the Hellenic Artificial Intelligence Society (EETN,