Singapore is moving to put artificial intelligence into practical financial services applications, with new initiatives targeting scam detection, AI-powered payments, and financial-sector talent. DBS, OCBC, Standard Chartered, and UOB are among the program's launch partners.

Singapore’s Monetary Authority of Singapore (MAS) and the Future of Finance Institute (FFI) have announced a series of initiatives to accelerate artificial intelligence adoption across the financial sector. The program will focus on strengthening financial crime prevention, enabling safe transactions by AI agents, and developing specialist AI capabilities.

Announced on 18 November 2026 at the Singapore FinTech Festival, the initiatives support Singapore’s National AI Missions for financial services. They bring together financial institutions, technology companies, research organizations, and public agencies to develop practical applications.

The program includes an AI Innovation Arena offering synthetic data for model development, an Agentic Money Sandbox for controlled financial transactions and new collaborations to advance speech recognition and AI talent.

Four major financial institutions have also joined FFI as launch partners: DBS, OCBC, Standard Chartered, and UOB.

The announcement signals a shift from AI experimentation towards shared infrastructure and practical testing across Singapore’s financial system.

AI Innovation Arena to Support Scam Detection

The first initiative will give technology firms and researchers access to synthetic financial data through the AI Innovation Arena.

FFI will work with MAS to provide developers with datasets that reflect real-world financial activity. These resources will help companies build and test AI solutions against industry challenges.

The initial focus will be financial crime prevention, particularly scam detection.

MAS will provide a synthetic dataset based on interbank transaction patterns and scam activity identified through an earlier proof-of-value exercise involving five banks. The dataset was developed with the Government Technology Agency of Singapore (GovTech).

Importantly, the dataset is synthetic rather than a public release of participating banks’ live customer transactions. It is designed to help developers train and evaluate AI and machine learning models.

Technology firms, AI research labs, solution providers, and research institutions can submit expressions of interest to access it.

This approach could help developers test detection models against more realistic transaction patterns. It may also support solutions that identify suspicious activity across financial institutions.

However, the initiative does not yet establish that AI models have improved scam detection. Their performance will need to be assessed through testing.

Over time, the AI Innovation Arena could expand to support other national AI missions and additional industry challenges.

Agentic Money Sandbox to Test AI-Driven Payments

A second initiative will address one of the emerging challenges in financial technology: allowing AI agents to move money safely.

FFI will establish an Agentic Money Sandbox where financial institutions can develop and test how AI agents transact using digital money.

The sandbox will operate within defined authority, transaction limits, and safeguards. Its initial use cases will include agent-initiated payments and treasury management across participating financial institutions.

This is an important step as financial institutions explore agentic AI systems that can perform tasks with less direct human intervention.

For example, an AI agent could initiate an authorized payment or support corporate treasury operations. However, these activities require clear controls over what an agent can do, which funds it can access, and when it must seek further authorization.

The sandbox will allow participants to examine these arrangements before deploying such applications in live environments.

The focus is not simply on whether an AI agent can execute a transaction. It is also on whether financial institutions can verify its authority and keep activity within approved limits.

This distinction will become increasingly important as AI agents take on more responsibilities in banking and payments.

MAS and A*STAR Advance Financial AI Capabilities

The third area focuses on strengthening AI capabilities for financial services.

FFI will work with A*STAR’s Institute of Advanced Intelligence and Computing (IAIC) to advance locally grounded speech-to-text technology into live pilots.

The work builds on an initiative announced by MAS at the Singapore FinTech Festival in 2025.

A key objective is to develop speech recognition that can transcribe Singlish and a mix of commonly spoken languages and dialects. This could make AI tools more useful in Singapore’s multilingual financial environment.

One potential application is real-time compliance monitoring for operational risk management.

Speech-to-text technology could help financial institutions analyze conversations and identify issues that require further review. The collaboration will also explore other capabilities, including generative AI guardrails.

These safeguards are important as banks and financial institutions expand their use of generative AI. They can help organizations manage risks associated with inaccurate outputs, inappropriate responses, and the handling of sensitive information.

FFI will also partner with AI Singapore (AISG) to develop AI talent for the financial sector. Potential activities include apprenticeship programs that connect financial institutions with AISG’s expertise and wider AI network.

DBS, OCBC, Standard Chartered, and UOB join FFI.

MAS and FFI have named four financial institutions as FFI’s launch partners.

They are:

  • DBS
  • OCBC
  • Standard Chartered
  • UOB

These institutions will work alongside FFI’s technology and research partners. Their participation is intended to bring financial-sector expertise and practical industry requirements into the development of AI solutions.

The collaboration model is significant because many financial challenges extend beyond individual institutions.

Scams, for example, can involve transactions moving across several banks. Shared datasets and coordinated testing could help developers examine patterns that are difficult to identify from a single institution’s perspective.

Similarly, safe agentic transactions will require financial institutions to consider common approaches to authority, safeguards, and interoperability.

The four launch partners should not be confused with the five banks involved in the earlier scam-detection proof-of-value exercise. MAS has not identified those five institutions in the announcement.

Singapore Builds a Shared Framework for AI in Finance

The new initiatives support three National AI Missions for financial services: strengthening financial crime prevention, improving access to financing for businesses, and enabling safe agentic transactions.

They also reflect a broader policy approach. Rather than leaving each financial institution to develop its own AI capabilities independently, Singapore is encouraging collaboration between banks, researchers, technology providers, and public agencies.

This could reduce duplicated development work and help promising solutions reach more institutions.

However, several challenges remain. AI models must perform reliably across different financial environments. Institutions must also manage data protection, operational risks, and accountability when automated systems influence financial decisions.

Agentic payments introduce another layer of complexity. Financial institutions will need clear rules for agent identity, permissions, transaction limits, and oversight.

The new sandboxes and development programs provide a setting to examine these issues. They do not, however, mean that the technologies are already approved for unrestricted use.

The next test will be whether these shared initiatives can produce reliable solutions that financial institutions can deploy at scale.

For Singapore, the ambition goes beyond adopting AI tools. It is to build the data, safeguards, talent, and industry collaboration needed to make AI a trusted part of financial services.

Source: Monetary Authority of Singapore (MAS), 18 November 2026; Singapore FinTech Festival 2026 ministerial keynote.