SMS Tan Kiat How’s Keynote Address at the Asia New Vision Forum Innovation 2026

Distinguished guests, ladies and gentlemen,

1. Good afternoon. It is a pleasure to join you at the Asia New Vision Forum. Thank you for inviting me to speak on Singapore’s National AI Strategy.

2. We meet at a remarkable moment for innovation. AI is becoming more capable. Autonomous systems are moving into the physical world. Advances in biology are changing how medicines are discovered and produced. AI agents are beginning not only to answer questions, but also to act, negotiate, and transact.

3. Technology is also increasingly part of the strategic conversation among the world's largest economies. The recent meeting between President Trump and President Xi is another reminder of that. The two sides agreed to establish a dialogue on AI, including its risks and benefits. So, AI is no longer just a technology matter, it is becoming an integral part of the economy, and certainly when the leaders of the world’s two largest economies meet.

4. But businesses cannot wait for every geopolitical question to be settled. They still have to innovate, invest, serve customers, enter new markets, and make decisions about where the next decade of growth will come from.

5. So, for the founders, CEOs, investors, and policymakers here – and for many companies watching from China – there is a more immediate question: How do we turn technology into value, and value into scalable businesses?

6. In Chinese, we often talk about “赛道”, or the racetrack. Much of the attention today is on one race: who can build the most capable AI. Better models. More compute. More advanced chips. More energy. More talent. That race matters, and it will continue.

7. But for most companies, there is another race that matters just as much: who can turn rapidly-advancing technology into real economic value – repeatedly, reliably, and at scale?

8. Give two companies exactly the same AI model. Will they create the same value? Almost certainly not. One may have better data. Better workflows. Stronger engineering. More capable people. Better management. Better partners. And a greater willingness to redesign how work gets done. The technology may be the same. The system around it is not.

9. That is why the next race is not only about making technology more capable. It is about how much more capable technology can make our companies, our workers, and our economies.

10. And for many Chinese enterprises, there is a further question. You may already have strong technology, deep engineering capability, and large domestic scale. But how do those capabilities travel? How do they work with different customers, regulations, standards, partners, and operating environments? And how can Singapore play a useful role, for many of these companies grappling with these issues?

11. For Singapore, the answer starts with a simple principle, I call it “看准赛道”  – choose the right race. We are not the world's largest market. We will not compete simply on scale. But we can compete on how effectively we can bring different capabilities together – and how quickly we turn innovation into impact.

12. For investors, this changes the question too. It is no longer enough to ask who has the best model, the best chip, or the most impressive demonstration. The harder question is: who can convert technology into a trusted and repeatable business?

13. For Singapore, I see three capabilities becoming especially important: conversion, assurance, and interoperability. Conversion helps you to build. Assurance helps you to prove. Interoperability helps you to scale.

14. These are not abstract policy ideas. These are economic and system capabilities. They determine whether a good technology becomes a successful business - and whether a successful business can become a scalable one.

15. Let me start with conversion. A strong technology is not yet a successful product. It has to solve a real problem. It has to fit into workflows. Users have to trust it. The economics of it have to work. And the technology, business model, operating model, and regulatory model have to come together.

16. Take autonomous vehicles for example. WeRide has developed sophisticated autonomous-driving technology in China. In Singapore, it is working with Grab to bring that technology into a local and unique operating environment.

17. WeRide brings the technology. Grab brings deep regional operating experience. Singapore brings real roads, commuters, a demanding safety regime, and regulators prepared to work through the practical questions of deployment.

18. In Punggol, their vehicles have gone through safety assessments and extensive road testing before public deployment, while our Land Transport Authority continues to monitor safety and performance. So, Singapore provides more than a test site; it provides a place where companies can learn what real-world deployment actually requires.

19. The question is not only whether the technology works. It is whether the whole system around the technology works. Can it operate safely? Can it integrate with transport networks? Can users understand it? Can regulators oversee it? Can operators build a viable service around it?

20. Soon, the system will become even more interesting. Waymo is also coming to Singapore, with a phased plan towards commercial autonomous ride-hailing here.

21. We should not assume that the world will converge on a single autonomous-driving technology stack. In some markets, national security, data, and regulatory considerations may constrain which technologies are deployed. Different ecosystems will persist.

22. But in much of the world, technologies from different ecosystems will still have to coexist in real life. They do not need the same software. They do not need the same sensors. They do not need the same business model. But they will share the same roads.

23. They must meet common expectations of safety. They must understand the same road rules. They must interact safely with human drivers, buses, cyclists, pedestrians,  and eventually with one another.

24. Different technologies. Common rules. Shared roads. That is interoperability in the physical world.

25. But being deployable is not enough. Before innovation can scale, people must trust it. Customers. Partners. Investors. Regulators. And for businesses, assurance is not an abstract governance concept.

26. Trust is part of market access. If a regulator does not trust the system, it may not be deployed. If a customer does not trust it, it will not be adopted. If a partner cannot rely on it, they will not integrate. If investors cannot understand the risk, capital becomes more expensive.

27. So, assurance is not what comes after innovation. Increasingly, it is how innovation reaches the market. We see it with AI: as systems become more capable and more autonomous, companies need credible ways to evaluate how they behave, where the limits are, and what happens when something goes wrong.

28. And we see it in biopharma. WuXi XDC established its first overseas manufacturing facility in Singapore's Tuas Biomedical Park. In roughly two and a half years, the facility moved from announcement to GMP release. Singapore is now part of WuXi XDC's broader global manufacturing network - complementing its capabilities in China, diversifying production and supporting global customers.

29. For a global customer, a manufacturing facility is valuable not simply because it can produce. It is valuable because there is confidence in the quality of what it produces, the processes behind it, and the standards that govern it. A lead market is therefore not simply somewhere innovation can be tried. It is somewhere innovation can earn trust. And in a world where technologies are becoming more powerful, that trust itself becomes an economic asset.

30. In many sectors, capital follows confidence. The more clearly the risk can be understood, tested, and governed, the easier it becomes for customers to adopt, partners to integrate, and investors to commit.

31. The same principle increasingly applies to the digital economy. For companies going global, internationalisation is not simply selling the same product in more places. It means operating across different regulations, payment networks, standards, identities, data requirements, and customer expectations. Those differences will not disappear.

32. So, the practical question is not whether every system can become the same. It is whether different systems can still work together.

33. We can already see this in agentic commerce. Imagine an AI agent arriving at a payment network and saying: I am authorised to make this purchase on behalf of this customer. How does another system know that is true? Who is the agent? Who authorised it? What is it allowed to do? What limits apply? And who is accountable if something goes wrong? These are real practical questions that businesses are looking into.

34. Ant International, Mastercard, and Visa are working on a Know-Your-Agent interoperability framework in Singapore. They are not trying to become one network. They are developing common approaches for agent onboarding and identification across networks, while preserving each network's own verification and decision-making. Interoperability does not require uniformity. Different systems. Common trust. Wider markets.

35. I see this in action on a regular basis. I have the privilege of co-chairing the Singapore-China Digital Policy Dialogue (DPD) with China's National Data Administration. At our inaugural Dialogue a few years ago, I proposed a simple approach: “一心务实” - stay practical, stay focused, and work on concrete issues that matter to businesses.

36. Because businesses do not experience digital policy as an abstract debate. They experience it when data cannot move. When a document cannot be recognised. When two regulatory systems use different concepts and companies have to pay lawyers and consultants simply to translate them. When compliance uncertainty delays a product launch or makes an SME decide that international expansion is simply too difficult.

37. So, our DPD has focused on practical issues: trusted cross-border data flows, AI governance, digital trade, digital credentials, and policy innovations that reduce friction between our markets. We have worked on guidance that helps businesses navigate different data requirements. We have worked on verifiable digital credentials that can make cross-border documents faster and easier to authenticate.

38. These sound technical. But to a business they mean something very concrete: lower cost, greater speed, and more certainty. The objective is not making our systems identical, it is to make it easier for businesses to operate across them.

39. That brings me to another phrase I used in our Dialogue: “和而不同” – harmonious despite our differences. Singapore and China do not have identical digital ecosystems – we do not need to. What matters is whether we understand the differences, build trust across them, and create practical bridges between them.

40. If “一心务实” describes how we work, and “和而不同” describes what we are trying to achieve. And the commercial consequence is important: interoperability reduces the cost of doing business across borders. Fragmentation is not just inconvenience. It is an economic tax. Large companies may be able to absorb that tax. Smaller companies may simply stay home.

41. This brings me to how we see Southeast Asia. I have spoken to many of you who are interested in the market. Singapore itself is a small market, but we sit within a much larger and highly diverse region. Different languages. Different consumer behaviours. Different regulations. Different stages of development. That diversity can make internationalisation harder. But it also makes the ability to work across systems more valuable.

42. Singapore can be a useful first market where companies work with sophisticated customer base, regulators, and regional partners – and build the capabilities needed to operate more widely. Singapore can be a lead market. ASEAN is the scale opportunity — and beyond ASEAN lies the global opportunity.

43. And scale is not only a growth story. It is an investability story. A company that can adapt across markets, regulations, and customer contexts is more resilient and potentially more valuable. The proposition is complementary: strong capabilities at home, a demanding lead market in Singapore, then scale across a diverse region.

44. ASEAN itself is putting in place more of the architecture to support this. The region has concluded negotiations on the ASEAN Digital Economy Framework Agreement with the aim of building a more integrated, secure, and interoperable regional digital economy.

45. Singapore has pursued the same philosophy internationally such as Digital Economy Partnership Agreement, which promotes trusted data flows, digital trade, and interoperability between different regimes. The details can become very technical. But the business objective is simple: Make it easier for successful solutions to cross borders.

46. That matters to an investor assessing whether a company can scale. It matters to an entrepreneur deciding where to launch. And it matters to an established company deciding where to build its next regional capability.

47. So, putting the three capabilities together – conversion turns technology into something that works, assurance turns innovation into something people can trust, and interoperability turns a successful deployment into something that can travel.

48. This is why Singapore’s ambition is to be more than a testbed. A testbed asks whether the technology can work here. A successful pilot proves possibility. A repeatable deployment proves a business. A lead market asks a more difficult question: can the technology create value, earn trust, and scale beyond here?

49. So, as technology becomes more widely available and AI becomes more pervasive, the difficult part increasingly moves downstream: from invention to integration, from prototype to deployment, and from one market to many. The next generation of global hubs will be judged not by what they attract, but by what they help companies build, prove, and scale.

50. That is the opportunity I see for Singapore: not merely to host innovation, but to help make innovation investable, trustworthy, and exportable. For companies in China looking outward, Singapore offers complementary strengths - not replacing what you have built at home, but helping those capabilities travel.

51. WeRide illustrates conversion. WuXi XDC illustrates assurance. Ant International's work with Mastercard and Visa illustrates interoperability. And the DPD shows how governments can reduce the policy and regulatory frictions that sit behind all three. Strong capabilities at home. A lead market in Singapore. Scale across ASEAN and beyond. That, in a nutshell, is how I’d describe Singapore’s National AI Strategy, to play a useful role in the next phase of AI.

52. If I can leave the audience here with four Chinese phrases, the first  phrase is “看准赛道”, the second is “和而不同”, the third is “一心务实”, and the fourth is “从‘新’出发”.

53. 在结束之前，我想用几句话，用中文概括今天的想法。

54. 第一，看准赛道，把科技转化为价值。不只是一味追求规模，而是看清自己的优势，把创新真正转化为生产力、竞争力和增长。

55. 第二，和而不同，以互信促互通。不同的技术、制度和市场不必完全一样，但可以在尊重差异的基础上建立信任、减少摩擦、加强连接。

56. 第三， 一心务实，以合作解难题。聚焦企业真正关心的问题：数据流动、规则衔接、标准互认、跨境经营，一件一件去解决。

57. 第四，从‘新’出发，从新加坡走向世界。在新加坡开发，在新加坡验证和部署；从新加坡出发，走向东南亚，走向全球。

58. 看准赛道, 和而不同, 一心务实, 从‘新’出发。These are the four phrases that encapsulate the key thrusts of my speech.

59. Ladies and gentlemen, let me conclude, those four phrases are more than a summary. They describe the role Singapore can play in a world that may become more technologically diverse, but will remain deeply connected.

60. The technology ecosystems of the future may not converge. But commerce still needs to flow. Vehicles still need to share the same roads. AI agents will need to transact across networks. Data and digital documents will need to cross borders. And companies will need places where trust can be built across differences.

61. Singapore can be such a place. Not because we are the biggest market, or because we determine where the technology frontier moves. But because we can be a useful market - one that helps turn frontier capability into trusted and scalable solutions.

62. For a small country, relevance does not have to come from scale. It can come from being useful. A place where different technologies can meet real-world standards. Where trust can be earned. Where companies can learn what it takes to operate across markets. And where what works can travel.

63. And that is also an invitation - to our innovators, investors, and companies looking outward across Asia. To choose the right race. To work across differences. To build trust and connectivity. And to use Singapore as a place from which good ideas can reach a much wider audience around the world.

64. So, to companies looking outward from China and across Asia, my invitation is simple: Bring your capabilities. Build with us. Prove what works here. Then take it from Singapore to ASEAN and the world. Build here. Prove here. Scale from here.

65. Thank you very much.