Transcript of Acting Minister for Manpower And Senior Minister of State, Ministry of Digital Development and Information Jasmin Lau's Address at AI Tech Festival

Introduction

1. A very warm welcome to everyone.

2. We have all heard a lot about AI. AI conferences, courses, and new tools are everywhere.

3. I think it is fair to say that some of us are feeling a little fatigued. Perhaps some of you are also thinking: “I have a job to do. What does all this AI hype mean for me?”

4. For most public officers, AI is not our job. We still have policies to develop, cases to handle, services to deliver, people to manage and Singaporeans to serve.

5. So, when we set up the Institute of Digital Government, or IDG, with CSC, we started with a simple question: How do we help every public officer use technology in a way that is genuinely useful to their work?

6. The work, purpose and mission still come first, and technology is the tool. Our aim is not to teach digital, data and AI for their own sake. It is to help officers do their jobs better and Government to deliver better outcomes for our citizens.

7. To do this, we decided to start with leaders. The leaders have to be role models before we can help every officer learn through real work, support those who are ready to go further, and spread what works.

First, start with leaders

8. We deliberately started with our Public Service leaders.

9. Nearly 150 Permanent Secretaries, Deputy Secretaries and heads of agencies started building their own digital capabilities two years ago. More recently, they rolled up their sleeves and tried building with AI themselves, from developing morning briefs to creating personal agents.

10. Why start with leaders?

11. Because leaders shape the conditions in which the rest of us work.

12. By trying these digital tools themselves, leaders can better understand what is useful, what the frustrations and risks are, ask better questions, and give their teams the confidence and space to experiment.

13. NLB Chief Executive Officer Melissa Tam worked with archivists to understand their pain points and vibe-coded an app to reduce the manual tracking of declassification requests and speed up responses to the public. Through the process, she realised that the bigger opportunity was to rethink the workflow itself, rather than simply layer technology onto it.

14. Leaders do not need to become technical experts. But they need to understand enough to lead change and protect time for their teams to learn and try.

15. I often meet business leaders and they tell me how AI is starting to feel like a KPI. They shared that they must first think about the problem they are trying to solve before applying AI. That is the mindset we need to build among our leaders.

16. We have services to deliver and problems to solve. If AI can help us, that’s great. But it doesn’t mean you have to use AI to solve it if there are other ways to get the job done. Nobody should feel pressured to have to use AI every day if it does not help you.

17. What we are doing here is to make sure you have the tools, the knowledge of how they can be used, and then you make the judgement whether it is useful for you.

18. To help, IDG has curated a 12-week learning plan where teams set aside an hour a week to build digital capabilities and apply what they have learnt together. This has been accessed more than 800 times.

Second, learn through real work

19. The second part of our approach is about all of us: Learning through real work.

20. The starting point is simple: before asking, “What can AI do?”, ask: “What are we trying to do better?”

21. Maybe officers spend too much time searching for documents or a service is difficult for citizens or businesses to navigate. Sometimes the problem lies in the process, the data, or how the work is organised.

22. That is why IDG starts with product thinking: Understand the problem first, then decide whether technology can help.

23. Every officer will need a common foundation: how to understand user needs and problems, use technology responsibly, and exercise judgement when working with AI.

24. From there, learning should become practical. Learning should follow the work, not the other way around.

25. Our goal is that by 2028, more than 150,000 public officers will be able to use technology effectively in everyday tasks and in the work of their profession.

26. IDG will provide bite-sized guides for different levels of fluency, with idg.gov.sg ↗ (opens in new tab) pointing you to relevant resources.

27. Start with one task that takes too long and start experimenting. Not everyone needs to become a builder, but everyone should be able to find something useful to them.

28. Confidence comes from working on a real problem, discovering what works and what does not, and improving from there; not sitting through presentations.

29. That is why we are expanding AI Sprints. These are exercises where officers bring real problems into an environment set up by IDG, and then work from problem to prototype.

30. Since May, more than 1,000 officers across the Government have participated in the pilots. In one Sprint involving over 330 HR leaders, more than 80 per cent felt more confident using AI afterwards.

31. Sometimes the answer is a simple prompt. Sometimes it is an agent or a redesigned workflow. And sometimes you discover that AI is not the answer at all. That is still useful learning, and I encourage all of you to first think: “Can it be solved by humans before thinking, let's reach for AI?”

32. The point is not to use AI. The point is to solve something real.

Third, help those who are ready to go further

33. The third part of our approach is to help our more confident users go further.

34. Some of you will move from using AI for individual tasks to building solutions and redesigning work.

35. Take Matthew, an architect at HDB. He is not a coder. His job is designing buildings, not bots.

36. Matthew used to spend hours searching through guidelines from HDB, BCA and SCDF to work out which requirements applied to each project. So, he built an AI assistant to help him.

37. What used to take hours can now take minutes, giving him more time for the work that requires his professional expertise.

38. Matthew did not start by asking, “How do I use AI?” He started with a problem in his own job and built something useful to solve it.

39. That is what we want to enable more officers to do.

Fourth, look out for those who need support

40. While there are confident users, and there are also less confident users. There are officers among all of us who feel a little bit more worried about AI.

41. They see the harms, and maybe they are worried about how the data centres affect the environment. There will be those among us who feel uneasy when they hear words like prompt engineering and agent orchestration.

42. For those of you who are comfortable, I ask that you look out for those who are less comfortable.

43. Don't let AI drive a fissure in our public service. If you are the confident one and you know of someone in your team who is a little bit more anxious about the tools and the technologies, spend a little bit more time with them.

44. Share with them your experiences, the benefits and the downsides, so that even the ones who feel a bit less confident are able to say, “Maybe I'll learn something about AI today. Even if I don't use it in my work, maybe I'll learn about its impact on society, its risks, and how to manage it better.”

Finally, spread what works

45. A good idea should not stop with one person or one team.

46. If someone, like Matthew, finds a better way of doing something, we want others to learn from it, adapt it and take it further.

47. That is why we are building a community of Champions of AI Transformation, or CATs: Officers who will drive workflow redesign in their teams and share what they learn with others.

48. Some of the best learning comes from colleagues who have already tried something themselves.

49. Every attempt to build something or to do a technology project requires significant investment of time, energy, and breaking down barriers. So, where possible, share your failures and your lessons.

50. This afternoon’s breakout sessions are opportunities to learn directly from people doing similar work: Ask them what worked, what did not, and how you can adapt.

51. That is how one useful idea can travel beyond one team and start to change how the Government works.

What success looks like

52. We should judge IDG’s success not by the number of courses we run or AI products we deploy, but by whether technology helps officers work better and deliver better services to citizens and businesses.

53. Around 70% of public officers are already using AI regularly. The next step is bigger than adopting a new tool: It is changing how work gets done.

54. AI can take away some repetitive work, giving officers more time for work that requires judgement, context and care.

55. I also believe that AI can free up some time for many public officers to spend a bit more time with your families.

56. But technology must remain a means, not become the mission.

57. Because every workflow we change has a person on the other side: an elderly resident trying to redeem a voucher, a young mother navigating healthcare forms late at night, or a small business owner waiting for an approval for a grant.

58. That is ultimately what IDG is here to do: To give our public officers tools to help build a government that is simpler, faster and more human for the people we serve.

Conclusion

59. Before I end, I encourage you to treat today as your own first Sprint.

60. Attend a workshop or plenary. Ask questions. Try a tool. Learn a skill. But most importantly, bring it back to your workplace on Monday and try it on a real problem.

61. Do not do it simply because everyone is talking about AI. Do it because it helps you work better and serve Singaporeans better.

62. Share what you have learnt, including what did not work.

63. Because that is how change happens: we learn what works, build on it, and take it further together.

64. That is how thousands of public officers can learn, build and serve Singapore together.

65. Thank you, and please enjoy the festival.