Global: MAS Deploys AI Initiative Using Live Bank Data to Strengthen Fraud Detection

The Monetary Authority of Singapore is collaborating with the Government Technology Agency of Singapore and the Singapore Police Force on a new initiative aimed at strengthening the country’s fight against financial crime through artificial intelligence and machine learning technologies.

The project is focused on enhancing scam detection capabilities by leveraging AI and machine learning models trained on transaction data drawn from multiple financial institutions.

As part of the initiative, the central bank is currently conducting a Proof-of-Value (POV) exercise involving five participating banks. The exercise is designed to assess how AI and machine learning models can be used proactively to identify suspicious transactions and detect scam-related activities before losses occur.

According to the MAS, the project will utilise historical transaction data and actual bank account numbers from participating institutions to train and evaluate the effectiveness of the models in identifying fraud patterns and financial crime risks.

To address concerns around privacy and data protection, the regulator stated that participating banks have been provided with a secure data-sharing environment governed by strict policies and protocols aimed at safeguarding customer information.

The framework includes the hashing of account numbers, ensuring that only the originating financial institution can identify its customers while allowing anonymised data to be analysed collectively for fraud detection purposes.

The Monetary Authority of Singapore noted that the initiative represents an important step toward deeper collaboration between regulators, law enforcement agencies, and financial institutions in combating increasingly sophisticated financial crimes.

According to the regulator, the use of AI and machine learning technologies is intended to complement existing fraud prevention systems already deployed by banks, while improving the speed and accuracy of scam detection across the financial ecosystem.

Industry analysts say the initiative reflects a growing global trend among regulators and financial institutions toward adopting advanced technologies to strengthen anti-money laundering controls, fraud monitoring, and transaction surveillance frameworks.

The MAS further indicated that insights from the Proof-of-Value exercise could pave the way for broader deployment of more advanced AI-driven monitoring systems across the banking sector.

Future phases of the initiative may involve expanding the datasets used for model training, increasing the sophistication of machine learning tools, and applying the technology to a wider range of financial crime and compliance use cases.

The regulator added that the long-term objective is to build a more resilient and secure financial system capable of responding effectively to evolving criminal threats in the digital economy.