AI tax system helps Malta collect over €1.6 billion in arrears
VALLETTA (MALTA) (ITALPRESS/MNA) – Malta’s artificial intelligence-powered tax monitoring system has helped the authorities recover more than €1.6 billion in tax arrears in less than four years, figures provided by the Malta Tax and Customs Administration (MTCA) show.
Since its introduction in 2023, the system has significantly improved the administration’s ability to identify tax evaders and late payers, with €455.4 million recovered by September 2026 alone.
The system, known as SAS, helped collect €374.8 million in arrears in its first year, followed by €431.3 million in 2024 and €419.7 million in 2025.
VAT arrears accounted for a substantial share of the recoveries. The MTCA collected €196.9 million in 2023, €183.6 million in 2024 and €165.5 million in 2025. By September 2026, VAT arrears recovered had already reached €202 million.
MTCA auditors have developed more than 100 risk rules within the system, allowing it to identify discrepancies between income tax and VAT declarations, late submissions and returns containing missing information.
SAS also categorises taxpayers according to their payment behaviour, distinguishing between those who pay on time, those who pay late and those who fail to pay altogether.
Last month, Finance Minister Clyde Caruana announced that several hundred people would receive letters seeking explanations for discrepancies between their declared income, taxes paid and significant purchases.
The minister said the technology was making tax evasion easier to detect.
The technology forms part of a wider digitalisation programme aimed at transforming the MTCA from a largely reactive administration into a more proactive enforcement authority.
Looking ahead, the administration is preparing to issue a tender for technology capable of monitoring VAT reporting in real time.
Similar systems introduced in other countries have helped reduce the VAT gap by between 15% and 20%, Caruana said.
– Photo Ipa Agency –
(ITALPRESS).
Source: medNews
