Speeches
Published Date: 19 November 2026

"Advancing the Future of Finance: From Foundation to Scale" - Keynote Address by Mr Chia Der Jiun, Managing Director, Monetary Authority of Singapore, at the Singapore FinTech Festival 2026 on 19 November 2026

1. Good morning, and welcome to Day Two of the Singapore FinTech Festival (SFF).

2. Two transformative forces in finance continue to capture our attention and imagination.

3. They could deliver greater efficiency, lower costs at higher speed. They could remake how financial institutions and markets organise and function. They are reshaping how individuals, corporates and finance interact and transact.

4. I am referring to Tokenisation and AI. I have spoken about both in previous years.

5. Today I want to update you about how we see progress in these technologies, how we are thinking about them, and what our initiatives and collaboration are.

6. I will speak first about Tokenisation, and then AI in the second part.

Tokenisation

7. Last year at SFF 2025, I said that tokenisation had moved beyond experiments but had not achieved escape velocity.

8. Since then, there has been further progress globally. Large banks are offering tokenised deposits to corporate customers for treasury management. A growing range of financial assets have been tokenised, including bonds and repos, equities, precious metals and money market funds.

9. Nevertheless, the value of tokenised financial assets remains small relative to that of conventional markets.

10. To scale, three things are needed:

a) Commercially compelling use cases for tokenised financial assets;
b) Trusted forms of digital money;
c) And interoperable infrastructure to support settlement.

Tokenised Financial Assets

11. Let me update on tokenised financial assets first.

12. There is interest from our industry to explore collateral management as a use case.

13. Collateral sits at the centre of derivatives, secured funding and liquidity management.

14. Tokenisation can make the movement and posting of collateral more efficient, quicker and at lower cost.

15. Tokenisation of collateral and financial assets more broadly changes the form factor and the rails on which ownership is transferred. It does not change the economic substance of the underlying financial asset, whether it is cash, bonds or money market funds.

16. This is fundamentally different than crypto assets, which do not represent any underlying asset or value. MAS’ consistent view has been that crypto assets are not suitable for retail investors, while tokenised financial assets do have benefits for wholesale markets.

17. To scale tokenised collateral management, three key building blocks are needed.

18. The first building block is legal and regulatory clarity.

19. We are updating the Guide on Tokenisation of Capital Markets Products to further clarify the regulatory and risk management requirements.

20. Through the Singapore Academy of Law, we are engaging further with the legal profession to clarify issues relating to the tokenisation of capital markets products, especially under private law.

21. The second building block is high-quality collateral assets. MAS has already tokenised MAS Bills – $100m was issued and settled against Singapore Dollar CBDC in November. We will work with the industry to onboard and tokenise other commonly used collateral. This could include widely used foreign government securities and money market funds.

22. The third building block is for the whole ecosystem of industry participants to move together. Banks, brokers, dealers, clearing houses and custodians – all are needed to post, accept and manage tokenised collateral.

23. MAS and industry partners have published today the Guardian Tokenised Collateral Framework. It helps the industry with the operational requirements across the collateral life cycle, including eligibility, custody, valuation, return and enforcement.

24. MAS will continue to work with the industry to develop the practices and capabilities to manage tokenised collateral. How quickly this scales will be up to industry adoption.

Digital Money

25. Let me turn to digital money. Tokenised markets need trusted digital money to settle transactions.

26. Tokenised bank deposits, regulated stablecoins and wholesale CBDC can all play a role.

27. We have strengthened the foundations for each.

a) We have issued guidance on the tokenisation of bank deposits;
b) We are introducing a regulatory regime for stablecoins;
c) We have conducted live wholesale CBDC settlement for interbank payments and tokenised MAS Bills.

28. We have established a collaboration platform with the industry for digital money. This is the BLOOM platform. On the platform, the industry is applying digital money to real use cases, and drawing lessons for operating practices. BLOOM brings together 24 industry members and policymakers from the region.

29. We published the BLOOM Field Guide today. It details practical considerations for deploying digital money in various use cases.

Trusted Settlement Infrastructure

30. As the use of tokenised assets and digital money grows, trusted settlement on CBDC will be needed. It underpins settlement finality and confidence, especially for large value settlement.

31. MAS has been testing CBDC settlement through the SGD Testnet. It has supported interbank settlement and MAS Bill issuance.

32. We will now reposition this as an operational utility. It will be renamed SG Digital Ledger. It will support settlement across institutions, tokenised assets, forms of digital money and ledgers.

33. The scope of take-up will be led by market demand.

Advancing Connectivity across Networks and Infrastructure

34. Tokenised assets will also need to transact across networks, infrastructure and borders. There are two parts to our work in advancing connectivity.

35. First, connectivity across networks and infrastructure. MAS, together with a group of global banks and policymakers, has collaborated on this challenge.

36. Through the Global Layer One initiative or GL1, we now have a set of standards for interoperable institutional-grade digital asset infrastructure. Five global financial institutions have now established the GL1 Organisation in Singapore. This independent, industry-led organisation will take this work forward, to strengthen assurance around digital asset infrastructure.

37. The second aspect of our connectivity work is to establish mechanisms for cross-border transactions and settlement. We are working with international partners on such approaches.

38. With the Bank of England and Bank of Thailand, MAS has tested a synchronised mechanism to settle across separate central bank systems.

39. With Deutsche Bundesbank, we have considered how separate systems can verify settlement conditions before funds are released.

40. These efforts help to build practical capabilities to connect settlement infrastructures across borders.

41. So this is our approach – collaborate with the industry to lay the foundations and prepare the building blocks, so that the industry can scale where there are commercially compelling use cases.

Agentic AI

42. Let me turn to the second part of my remarks, which will be about AI, and more specifically AI agents.

43. AI agents have taken significant steps into the mainstream this year. In a widely read post, Cloudflare has said that AI HTTP requests have already exceeded human traffic.

44. We are seeing agents used in a variety of tasks: coding, operations, research, customer service, and as personal assistants.

45. Consumers have a growing interest for agents to manage purchases and payments. Businesses see benefits for agents to manage cash, payments and liquidity.

46. As agents move into activities with financial consequences, authority, accountability and control become essential. Some payment services providers and financial institutions are building agent interfaces into their offerings, but many financial institutions have yet to establish terms or channels for agents to access financial services.

47. Some users are giving agents their login credentials to financial accounts. I would not advise giving personal agents such access, as there is insufficient assurance of safety.

48. There are some challenges here.

a) The financial institution does not know if this is an agent that is acting on the full authorisation and control of the customer;
b) Nor whether the agent is transacting within its mandate and instructions.
c) And if agents transact incorrectly or beyond their instruction, who bears responsibility?

49. We need a trusted protocol and channel for agents to access financial services. Earlier this year, we published our work on Safeguards for Agentic Finance at Runtime or SAFR. The lessons from the first phase of SAFR are relevant to how customer agents should transact safely with financial institutions.

50. What were the lessons?

a) First, agents need bounded autonomy: enough discretion to be useful, but kept within defined boundaries.
b) Second, the identity and authority of an agent to act needs to be clearly established.
c) Third, mandates need to be clear and specific: actions that are permissible, limits, conditions.
d) Fourth, controls and safeguards need to be proportionate to authority. The greater the discretion and impact of action given an agent, the stronger the assurance and controls needed.

Develop a common protocol for agent access

51. In the new phase of SAFR, we will facilitate the development of know your agent protocol for trusted information exchange between customer agents and financial institutions.

52. Information relating to:

a) Who the agent is and whom it represents;
b) What authority it has been given;
c) What mandate it operates under; and
d) What assurance exists about its provenance and reliability.

53. This gives financial institutions the information needed to determine the terms and scope permitted to an agent. With this information, financial institutions can establish a dedicated secure channel for agent access.

Agentic Money Sandbox

54. A protocol and channel for agents to transact with financial institutions is the first step.

55. We will also set up a Sandbox with the Future of Finance Institute. The sandbox will bring two technologies together – AI Agents and Programmable Money, and apply them to two use cases – corporate treasury management and payments services.

56. In the corporate treasury management use case:

a) Banks could allow a corporate agent to monitor liquidity positions and due payments, decide when and where funds should move, and use tokenised deposits to execute permitted treasury actions.
b) The corporate sets the mandate.
c) Participating banks set the controls and determine the terms for accepting each agent instruction.

57. In the payments use case:

a) Banks and payment service providers allow customer agents to execute payment instructions using digital money such as tokenised deposits or stablecoins.
b) Participating banks or payments providers will verify the agent’s identity and authority and apply risk controls before executing the transaction.

58. These two use cases, and others that we will collaborate on in the future, will provide important lessons about benefits and safeguards.

The Path Forward

59. The two initiatives that I have highlighted: (i) common protocols for agent transactions and (ii) the sandbox for agentic money, will be important building blocks for the future of agentic transactions.

60. The future we see is an interoperable ecosystem of trusted interactions between agents, financial institutions and digital money.

61. We will need clear identity, authority and mandate for each agent, and in each financial institution, we need terms, channels and controls to accept agent instructions.

62. We will collaborate with the industry to develop these common approaches and protocols.

Agentic Markets and Economy

63. I will now speak briefly about what a future agentic economy may look like, and some implications.

64. We are at an early stage of agentic use. But adoption has the potential to scale rapidly, and we should start thinking about the potential impact on financial institutions, business models on markets and the broad economy.

65. AI agents could have far-reaching consequences for economic and financial decisions and actions. AI and AI agents can lower the cost and increase the speed of search and comparison, optimisation and decisions, negotiation and execution.

66. The result of this efficiency and speed is that AI can sharply reduce information asymmetries, and reduce search and information costs in economic decisions. It could also enable more dynamic and continuous pricing in markets.

67. More specifically, for consumers, AI agents can:

a) Reduce search costs by comparing products, prices and terms across many providers;
b) Constantly negotiate on the consumer's behalf or monitor for price changes, potentially securing better prices or more tailored products.

68. For businesses, agents can:

a) Automate procurement, inventory management and other routine commercial decisions;
b) Search across a much wider pool of suppliers and continuously compare prices, delivery times and other terms.

69. The result may be a greater frequency and velocity of consumer and business decisions and transactions, which are more dynamically optimised to time, attributes and price. There could be more economic activity and transactions that would otherwise not have occurred.

70. Agent-to-agent interactions are a relatively new area of research and more work is needed in this emerging area. There could be dynamic markets, where prices and terms are determined continuously through agent-to-agent negotiations.

71. Emerging studies of agent interactions in simulated environments also point to some potential risks. For example, agent collusion could occur even in the absence of an instruction to collude, with results of higher prices or weaker competition.

72. In another study, there were suggestions that more capable agents could secure better prices and terms than less capable ones. This could raise concerns about unequal access to capable models.

73. We do not know how quickly agentic transactions will grow, but it is likely that the future will be a mix of the human economy and a new agent economy. Business models and practices will have to adapt and respond to the new speed and mode of agentic transactions.

74. At a basic level, while most businesses have an online presence, their customer interface and revenue model were designed for engagement with humans, not with agents. Businesses, financial institutions and financial infrastructure operators will have to adapt to serving both human customers and agents.

75. Agents have the potential to move money and allocate assets more quickly for smaller incentives. Financial institutions will have to strengthen their value proposition to customers and agents, and potentially consider whether limits are needed on agent-instructed transactions to manage risk.

76. Regulators will have to consider the implications for financial stability from agentic participation in financial markets and movement of money.

77. Thinking ahead about the potential future agent economy will help inform the policies and standards needed, so that we can harness the benefits of more efficient markets and expanded economic opportunity while minimising the potential risks.

78. Importantly, consumers and businesses need to retain meaningful control over the objectives, mandate and actions of their agents. Autonomy has to be bounded. Actions and outcomes have to come with accountability.

Conclusion

79. Let me conclude.

80. MAS’ approaches to tokenisation and AI have followed the same fundamental elements.

81. We provide regulatory clarity, establish governance responsibilities, avoid prescriptive rules, and collaborate with industry to develop standards and good practices.

82. We set up sandboxes and public-private platforms to work on use cases, draw lessons, and translate these into policies, practices, protocols or infrastructure that are helpful to industry adoption.

83. We promote interoperability and common standards, and seek to connect financial institutions, customers, networks and infrastructure.

84. The outcome we seek is an innovative and resilient financial sector that is able to harness transformative technologies for the benefit of financial institutions, businesses and consumers.

 

*****