Skip to main content

MAS to study shared scam detection system in early 2027

Five banks are pooling transaction data in a trial to test whether AI can identify patterns beyond the view of individual institutions.

By Maya Santos

Illustration of connected bank accounts and digital transaction monitoring
Photo: Fintech News Network

The Monetary Authority of Singapore (MAS) plans to start a study in early 2027 on the design of a Scam Detection Utility for the financial industry. The work will draw on a trial in which five banks are securely combining transaction data to test artificial intelligence.

Deputy Prime Minister and MAS Chairman Gan Kim Yong announced the study on 18 November at the Singapore Fintech Festival. It will examine governance and operating arrangements for the proposed utility, initially covering Singapore accounts and data.

Gan said scam proceeds can move between several institutions within minutes, leaving individual banks with only a partial view. The initiative aims to identify suspicious activity sooner and allow faster intervention to reduce losses. MAS announced the five-bank trial in May.

Separately, MAS and the Future of Finance Institute will make synthetic data available to developers through the AI Innovation Arena. Created with the Government Technology Agency of Singapore, it will reproduce patterns of interbank transactions and scam activity from the trial without disclosing the underlying records.

Technology companies, AI laboratories, solution providers and research institutions can express interest in using the dataset to build and test detection tools. The arena could subsequently offer datasets supporting other national AI missions.

Gan outlined three such missions for financial services: preventing financial crime, widening financing access, especially for SMEs, and allowing AI agents to conduct transactions safely.

Keep reading: Banking & Fintech

All Banking & Fintech stories

Latest in Finance

All latest