VAT Evasion

Value Added Tax (VAT) is a cornerstone of modern economies capturing revenue from every stage of production to consumption. The various systems around the world are designed to be self-policing with every layer in the supply chain responsible for collecting and remitting their slice of VAT to their Treasury’s coffers; but with the responsibility to collect and remit any taxes comes the temptation to evade them.

VAT is no exception, its complex buy-side/sell-side payments, exemptions and credits makes this tax particularly vulnerable to evasion and fraud. The VAT Gap (the difference between what is due and what is collected) runs into hundreds of billions of dollars worldwide which not only erodes government revenues but creates market distortion and undermines trust in a self-policing system.

Modern Tax authorities are investing millions in analytic tools, digital reporting frameworks and targeted enforcement strategies to combat evasion in their jurisdiction while modern bad actors , driven by potential huge profits from fraud, are coming up with ever more elaborate schemes to deceive, and so begins an evolving cat and mouse contest, each trying to outsmart the other in an eternal game-loop!

VAT Evasion falls into two broad categories. One that attempts to hide income to reduce the tax liability, and the second, that attempts to manipulate the quirks in the system to make a profit. The perpetrators range from small traders making cash-in-hand transactions to highly sophisticated cartels running complex schemes that de-fraud millions. Both are equally illegal and carry hefty penalties for offenders in every tax regime worldwide.

Common Retail Evasion

• Unreported Cash Sales: Businesses (common in hospitality and construction) accept cash payments "off the books" to avoid recording the sale and the associated VAT liability.
• Misclassification of Goods: Traders deliberately report standard-rated goods as zero-rated or exempt to reduce the amount of tax they must collect and pay.
• Under-Registration: A business intentionally keeps its reported turnover just below the legal threshold for VAT registration to avoid the system.
• Fake Invoicing (Invoice Mills): Shell companies are formed solely to issue fake invoices for services or goods never supplied. “Legitimate” (i.e. VAT registered) businesses then use these invoices to claim Input Tax Refunds for VAT that was never paid. ITC fraud in India is estimated to be 51% of its GST gap!

Import and On-line Schemes

• Import Value Manipulation: Importers falsely declare a much lower value for goods arriving from outside their tax jurisdiction to reduce the initial VAT due at the border.
• Cross-Border Online Fraud: Sellers operating from on-line marketplaces in one country fail to declare or pay VAT in the country where their customers are located, pocketing the tax collected from the sale.

Global High-Impact Schemes

• Missing Trader Intra-Community (MTIC) Fraud: This is the most prevalent form of organised VAT fraud worldwide. It is most used in trading blocs with free trade agreements between countries (like the EU where MTIC fraud is estimated at around €90Billion).
  • Acquisition Fraud: A trader imports goods VAT-free from another country, sells them domestically with VAT included, and then disappears without remitting the collected tax to the government.
  • Carousel Fraud: A more complex version where goods (often high-value, low-bulk items like mobile phones or computer chips) are imported and exported multiple times in a circular chain of companies. Each loop allows fraudsters to pocket collected VAT or claim fraudulent refunds.
• Contra-Trading: This is a classic example of bad actors evolving their strategies to evade enhanced government scrutiny. Fraudsters use two separate carousels, one legitimate and one fraudulent, to offset the VAT liabilities of one against the credits of the other, making the scheme much harder for authorities to detect.

The OECD’s March 2026 report confirms that artificial intelligence is now a core tool for tax authorities globally. 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. Rapid digitisation and integration together with these technologies transition tax enforcement from manual, reactive audits to proactive, real-time risk assessment.

The use of AI transforms the landscape for enforcement agencies who for so many years were outclassed and outnumbered by their better funded adversaries. Higher analytic techniques making their way into the everyday armouries of tax authorities include:

Detection of Fraud in Real-Time. AI systems process millions of transactions in seconds to identify discrepancies and anomalies such as input/output tax mismatches as they happen. When integrated with mandatory e-invoicing (e.g., Italy’s Fattura Elettronica or Spain’s SII), authorities can flag fraudulent supply chains immediately.

Pattern Recognition and Machine Learning. This branch of higher analytics excels at detecting the leading indicators of emerging fraud especially Carousel schemes with circular trading patterns and new business registrations across borders. Algorithms learn from each confirmed fraud case enhancing their ability to spot new and evolving tactics.

Predictive Analytics. Authorities now assign risk scores to taxpayers based on a range of criteria including delinquent history, industry benchmarks, behaviour and even importance to the economy. The resources of the agency are then focused on companies and taxpayers in higher risk categories dramatically improving their effectiveness when compared to the random less effective audits in the past.

External Data to Increase the Aperture of Analysis. Cross referencing tax filings with external data sources including social media to spot lifestyle inconsistencies and even satellite imagery to track truck movements to match whey bills or undeclared assets like swimming pools.

Network Analysis. Identifies links between shell companies, shared directors and employees or associations between cartel members.

Challenges. Despite these gains there remains a gap between policy and execution due to gaps in data and a shortage of forensic expertise to evaluate automated red flags.

Finally, detecting evasion isn’t enough. These cases require a successful prosecution to enable recovery of due taxes and to act as a deterrent for others. India’s prosecution rate for GST evasion is under 10% despite heavy automation while Serbia achieves a 90%+ prosecution rate for AI detected evasion. The difference is court admissible evidence that provides explainability to judges who are keen to protect taxpayer rights and vary of “black-box models”.

Around the world, several countries have achieved remarkable success in reducing indirect tax evasion through bold reforms, digital innovation, and strategic enforcement. These case studies offer powerful insights into what works, and what other nations can learn from their experience.

Here are some of my favourites:

Brazil was a pioneer in e-Invoicing. Brazil introduced its Nota Fiscal Eletrônica (NF e) system as far back as 2005, requiring businesses to issue digital invoices validated in real time by tax authorities before it reaches the consumer. This system dramatically reduced invoice fraud, improved transparency, and strengthened audit capabilities. Brazil’s experience showed the world that real time data can transform tax administration, making evasion significantly harder. E-Invoicing is now the gold standard in digitising VAT.

India was the first country to integrate registration, filing, payments, and invoice matching into a single digital platform with the introduction of the Goods and Services Tax Network (GSTN). The reconciliation of invoices has reduced opportunities for fraudulent Input Tax Credit (ITC) claims, which according to a Comptroller and Auditor General (CAG) report contributed over 50% of evasion at State level. The digitised records in GSTN also provided their enforcement agencies valuable insights into illegal trading patterns.

Once evasion is detected, the enforcement agency does not always slap a heavy-handed prosecution notice on the guilty party (except for large criminal cartels), instead they send a letter inviting them to re-file their VAT return with an accompanying data file intended to “help the taxpayer with their return” Once the taxpayers see the evidence held by the tax department they invariably end up settling their accounts knowing the game is up. This soft-handed approach has resulted in recovering INR 20,128 crore ($2.4Bn) from GST evasion cases in 2024!

Italy, with one of the largest VAT gaps in the EU, provides another compelling example. Italy implemented mandatory e invoicing for B2B and B2C transactions. The results were immediate: false invoicing networks were disrupted, compliance improved, and VAT revenues increased. Italy’s success demonstrates that comprehensive digitalisation (not partial or voluntary adoption) is necessary to achieve meaningful results.

In Portugal and Serbia, where cash transactions were rife, the introduction of fiscalized point-of-sale (PoS) terminals, certified invoicing software and real time reporting has helped reduce VAT evasion particularly in high risk sectors such as hospitality and retail where cash transactions were rife. These countries’ approach highlights the value of combining technology with targeted sector specific interventions.

South Korea offers a different but equally effective model. Its cash receipt system incentivises consumers to request receipts by offering tax deductions and lottery rewards. This behavioural approach has significantly reduced under reporting in cash heavy sectors. South Korea’s success illustrates that combating evasion is not only about enforcement, but also about influencing taxpayer behaviour.

These case studies share several common themes. First, digitalisation is essential. Whether through e invoicing, real time reporting, or integrated platforms, technology provides the visibility and control needed to detect and prevent evasion. Second, reforms must be comprehensive. Partial measures often leave loopholes that fraudsters can exploit. Third, behavioural strategies matter. Incentives, simplification, and taxpayer support can significantly improve voluntary compliance.

VAT evasion isn't just a victimless crime; it drains the public funds used for schools, hospitals, and infrastructure. As fraudsters get smarter, the combination of real-time digital reporting and AI-driven oversight is proving to be the most effective way to level the playing field for the enforcement agencies and honest businesses.

Countries that have successfully reduced indirect tax evasion show that progress is achievable, but it requires political will, investment in technology, and a commitment to continuous improvement. Their experiences offer a roadmap for others seeking to strengthen their tax systems and protect public revenue.

Sophisticated evasion tactics like fake invoices, under-reported sales, and circular trading schemes can quietly hurt your margins and trigger costly audits.

Contact us today to see how our AI-powered analytics can help you fight back: detecting anomalies in real time, revealing hidden fraud patterns, and delivering predictive risk insights — so you can safeguard your revenue and maintain strong VAT compliance.