MAS to Start 2027 Study on Industry-Wide Scam Detection Utility Built on Five-Bank AI Trial
Abstract:MAS will begin a design study in early 2027 for an industry-wide Scam Detection Utility, drawing on a five-bank AI trial that pools transaction data to catch scams crossing bank boundaries.

Singapore's central bank will begin a design study in early 2027 to work out how an industry-wide Scam Detection Utility would be governed and run, building on a live trial in which five banks pool transaction data to test whether artificial intelligence (AI) can spot scam patterns that no single lender can see alone. Deputy Prime Minister and Monetary Authority of Singapore (MAS) Chairman Gan Kim Yong announced the plan at the Singapore FinTech Festival on 18 November, according to Fintech News Singapore.
Why Scams Cross Bank Boundaries
The case for a shared utility rests on how quickly scam proceeds move. Gan said a victim's money can pass through several institutions within minutes, and that each bank seeing only its own part of the transaction chain leaves the system unable to contain the loss.
"Scams are not confined within the boundary of one bank," he said in his festival speech. "We will not be able to mitigate the risk if each institution sees only its part of the transaction chain."
The stated aim is to help financial institutions detect suspicious activity earlier and intervene faster to limit losses. Gan framed the approach as two tracks running together: connecting information across the financial system while strengthening detection inside each institution, so that suspicious activity is identified sooner and more money is stopped from leaving.
The five-bank trial, which MAS first announced in May, is described as a proof-of-value. The banks pool their transaction data securely, allowing AI models to pick up patterns that no single bank could see on its own. The trial remains ongoing, and its findings are meant to shape the eventual utility.
What the Design Study Will Cover
The study will set out how the proposed utility would be governed and operated, including its governance and operating model, before any wider build. The work will initially focus on accounts and data in Singapore.
Gan signalled that the scope could widen over time. While the first phase centres on domestic accounts and records, he noted that financial crime crosses national borders too, and that there may be opportunities to strengthen cooperation and learn from partners in other jurisdictions.
Alongside the study, MAS and the Future of Finance Institute (FFI) will give developers access to a synthetic dataset through an AI Innovation Arena. Developed with the Government Technology Agency of Singapore (GovTech), the dataset is designed to reflect interbank transaction patterns and scam activity from the five-bank trial without sharing the original records.
Technology firms, AI labs, solution providers and research institutions can register their interest in using the dataset to develop and test scam detection tools. The arena could later supply datasets for other national AI missions.
One of Three National AI Missions
Financial crime prevention is the first of three National AI Missions for Financial Services that Gan outlined at the festival. The other two focus on improving access to financing, particularly for small and medium-sized enterprises (SMEs), and on enabling AI agents to transact safely.
The missions sit under a National AI Council chaired by Prime Minister Lawrence Wong, established earlier this year to drive AI transformation in key sectors including financial services. Gan said the missions address challenges that span institutions, and that MAS will bring together financial institutions, technology companies, researchers and public agencies to build shared capabilities. Promising solutions will be tested against global best practices and standards, and those that work well taken towards wider deployment.
The FFI is cast as the delivery vehicle. With its partners, it will facilitate knowledge sharing, formulate guidelines and standards for deployment, and build practical resources for AI risk management, including agentic AI, meaning AI systems that can act on a user's behalf. It will also develop and test solutions for use across the sector.
On the SME financing mission, Gan said many lenders still assess businesses largely on past performance and repayment history, and secure loans against personal guarantees and collateral. In his view, these approaches may not fully capture the creditworthiness and future potential of SMEs, or of newer start-ups with limited track records or assets. He said richer data could let AI identify patterns traditional methods miss, supporting faster, better-informed lending decisions and potentially lowering borrowing costs, with AI technology companies invited to test solutions using realistic synthetic data.
The third mission looks further ahead, at AI agents that carry out transactions on behalf of individuals and companies. Gan said such agents could help manage cash more efficiently and make payments across institutions faster and more reliably, but that greater autonomy must come with clear responsibility and accountability. Agents would operate within defined authority and limits, with institutions remaining accountable for actions taken on their behalf. Trials will examine how agents interact with purpose bound money, a form of digital money restricted to specific uses, so that payments can be released automatically once conditions are met.
Trust, Oversight and the Wider Backdrop
Gan tied the pace of AI adoption to public confidence. He said clear standards and effective safeguards allow institutions to test and deploy new solutions responsibly, and give businesses and people confidence that their money and data are protected and institutions are accountable. Financial institutions, he added, must remain accountable with strong human oversight, and Singapore will invest in people so they can work effectively alongside AI.
The festival speech also touched on computing access. Gan noted that the US Department of Commerce has announced new rules on remote access to advanced semiconductors, under which companies will require licences to export or re-export advanced chips to five countries, including Singapore. He said Singapore is working with affected companies to assess the implications and engaging US authorities on the rules and on facilitating access for Singapore companies, while stressing that the city-state does not condone firms using their association with Singapore to circumvent or violate the laws of other countries.
On payments, Gan said the Phase 2 report for PayNow Generation 2 was released that day, setting out how payments will be made simpler, safer and more accessible. It includes a new SGQR+ switch intended to let Singapore residents and tourists pay at any SGQR+ label using their preferred QR application, along with other enhancements.
For readers following Singapore's regulatory direction, the scam detection plan marks a step from a closed bank trial towards a possible shared piece of national infrastructure for monitoring payment chains. The design study, due to start early next year, is where questions of data governance, operating responsibility and cross-border scope are expected to be worked through.
Finally, WikiFX would like to remind everyone that before using any US stock, futures, or forex brokers, remember to check the platform's regulatory information, trading environment, field inspection reports, user reviews, and overall rating on the WikiFX App. This will help you effectively avoid problematic scam platforms and rogue brokers, safeguarding your capital.

Disclaimer:
The views in this article only represent the author's personal views, and do not constitute investment advice on this platform. This platform does not guarantee the accuracy, completeness and timeliness of the information in the article, and will not be liable for any loss caused by the use of or reliance on the information in the article.










