Guest-of-Honour Speech by Minister for Law and Second Minister for Home Affairs, Edwin Tong SC,...

Ms Low Pei Lin and Ms Yvonne Tang,
President and Vice President of the Asian Patent Attorneys Association

Mr Gerald Koh,
Vice President of the International Association for the Protection of IP,
Singapore Group

Mr Soh Kar Liang,
President of the Institute of Singapore Trade Mark Agents

Ms Boo Yee Swan and Mr Jonathan Foong,
Co-Chairpersons of the Law Society of Singapore's IP Practice Committee

Mr Tan Kong Hwee,
Chief Executive of the Intellectual Property Office of Singapore (IPOS)

Introduction

1. Good afternoon, everyone.

2. It's a real pleasure to join you this afternoon. You have just heard an update from the IPOS Registries, and later this afternoon, you will delve deeper into two specific areas - trade mark examination and prosecution, and also patent examination and prosecution.

3. More importantly, why are we here today? I am told this is an inaugural session - first time that we are doing this. It is really an opportunity for us to engage in a deeper conversation as a community, which I think is increasingly going to be important. You all work in different spaces, and it is important that while we deepen our knowledge in each of the specific fields that you are in, that we also look at what is on the horizon and try to interact and interface with each other on that horizon.

4. Indeed, having had conversations as a community, practitioners have told us that it is a valuable opportunity for more direct engagement with IPOS, and for more open discussions on IPOS' practice directions.

5. In turn, IPOS values the opportunity to hear all your perspectives, the viewpoints from each of your specific areas, sharing our thinking on emerging developments, and how we can strengthen Singapore's IP ecosystem together.

6. So against that backdrop and that theme, I thought I would use my few minutes here to step back a little and reflect on some of the broader questions behind today's discussion topics on IP examination and prosecution, to ask ourselves: How has the work of IP examination evolved over time? If you are going to have ambitions to move forward collectively, then I think we need to understand how that work has evolved, what that transformation entails, and what we need to do to get to the next stop. And collectively, what are the skillsets and the tools that you will need to have to stay ahead, because, ultimately beyond moving fast, I think it is also a relative thing, that we need to be better than our competitors.

A quarter century of change

7. This year marks the 25th anniversary of IPOS - a useful milestone, not so much as a moment for celebration today, but really as a marker in time.

8. Because if we look back over these 25 years, what stands out is not only how IPOS has evolved as an institution, but fundamentally, and very critically for all of us here, how the fundamental nature of IP work has changed. We are dealing with very different assets today compared to 25 years ago.

9. In the earlier days, when IP functions sat within the Registry of Trade Marks and Patents under the Ministry of Law, examinations were a far more manual process. I suppose one could argue that any work 25 years ago would have been a far more manual process, but I think IP examination is, in particular.

10. Those who joined IPOS in those early years will remember the office at Bras Basah Road, and you will remember a room full of compactors where physical files were stored. I remember my office back in the days when I was in practice, where you got to wheel the dusty compactors open and then search manually.

11. Searches were very manual, very painstaking, and sometimes, hit and miss. Trade mark examiners worked through physical registers and records, to identify potential conflicts, often with very limited digital support, if any.

12. When the patent examination team was later established in 2013, the shift had already begun - from physical publications and journals to digital databases. A bit like how we used to do legal research - we used manual, noted-up cases to something that was more digital, where you can search across different domains and different jurisdictions. But even then, the nature of the work remained highly labour-intensive. Patent examiners would have to sift through multiple patent office systems, read large volumes of documents in soft copy form, to identify relevant prior art.

13. At its core, a significant part of the job was still about one thing - finding information - trying to trawl through, find, and sometimes match information.

14. That is reflected in one of the most frequently cited studies from the Japan Patent Office, which found that examiners spent approximately 30% of their time searching for prior art, and 10% of their time reviewing and understanding their own search results. In other words, about 40% of the examination time and process was devoted just to locating and then understanding and digesting prior art - before the substantive legal analysis even begins.

IP world today

15. Today, that world has changed significantly.

16. Digitalisation has made global IP databases instantly searchable. Machine translation has also broken down different language barriers. You can now instantly translate across different languages. Examiners can now access millions of records within seconds.

17. And increasingly, artificial intelligence (AI) is accelerating and powering this shift. AI is now being used not just to search for keywords, but also to understand meaning and context.

18. Instead of relying solely on exact search terms, modern systems can now identify conceptual similarity. You do not have to see exactly how it looks like - the colours, the design, but the concept can now be something that can be compared across different domains. By doing so, you link inventions or trade marks that are described differently, but actually operate on the same underlying idea. I think that is a game changer when it comes to protecting ideas.

19. In patent search, AI tools can now scan vast prior art repositories and surface documents that are technically different, even when terminology differs significantly. A search by a machine learning-based image recognition system, for example, may surface earlier disclosures framed around pattern detection or computer vision classification.

20. In trade marks, image recognition tools can compare device marks across different jurisdictions, detecting visual similarity in structure and composition, rather than just relying on traditional textual descriptions. So combined with classification tools, they can help to suggest relevant classes of goods and services.

21. But beyond search itself, AI is progressing this industry at a great pace. It is increasingly used to rank, cluster, and then summarise search results, helping examiners more quickly filter large volumes of information and focus on what is most relevant.

22. Indeed, we are already seeing tangible outcomes, as I am sure all of you who are using such devices will experience. For example, in Brazil, the national patent office partnered with Chemical Abstracts Service (CAS), a division of the American Chemical Society, to implement AI-assisted prior art search, and they reported up to 50% reduction in examination time, more than 75% reduction in search time for a large share of applications, and up to 80% reduction in backlog for the affected portfolio.

23. There are similar initiatives that are also being pursued by other IP offices around the world, including the European Patent Office and the Japan Patent Office, which have progressively integrated AI-enabled search, classification, and workflow tools. These help to support examiners in managing increasingly large prior art databases.

24. All of these developments and innovations are really only moving in one direction. It will just be more powerful, faster, and easier. You will be able to digest more information at a much faster speed. Technology is going to increasingly and fundamentally shape how we handle IP information, how we digest it, and how we can analyse it across different jurisdictions.

Challenges of scale and complexity

25. The reason AI is so transformative is really because of the scale. In almost every industry that is knowledge-based or document-intensive, as we are talking about here, the scale itself is being driven by something more fundamental, and that is the rising importance of IP. It is not just the size, scale, and the speed of the technology, but it also speaks to the nature of the business that we are involved in, the kind of IP products, and let me just elaborate on this.

26. We are now today living in an economy where intangible assets - ideas, innovations, software, data, know-how, information -form, today, the majority of enterprise value. It did not use to be the case; it used to be very brick and mortar. But today, intangible assets are the vast majority. In fact, in many leading companies, IP is no longer a supporting function; it is a core asset, because you are managing intangible assets as the majority of the enterprise value of any given company.

27. This shift is itself reflected in market structure. If you look at intangible assets, as a collective, they now account for roughly 92% of the entire market value of S&P 500 companies. This is up from around 70-75% about 25 years ago, and just 17% 50 years ago. So you can see the scale and pace of movement, and I dare say that we are not even at the bottom of the trajectory. We have not quite reached the end state of this evolution.

28. As IP becomes more and more valuable, it is also being created, filed, and commercialised at a much greater speed. The scale of transactions, the speed of transactions, and the need for us to do it in a seamless fashion across different domains will rise. This is reflected in sustained growth in both trade mark and patent applications globally, driven by digital platform business models, deep tech, AI innovation, as well as the globalisation of brands and technology.

29. The result is that a comprehensive IP search today that you might do, will draw on well over a hundred international databases, spanning hundreds of millions of technical records.

30. Therefore, in this domain, no examiner - no matter how experienced or how hard you might work - can really manually navigate this universe.

31. The challenge is no longer about sieving through the volume of information or having an eagle eye to match across different images and different domains, but how consistently we can interpret relevance across that volume, how consistent your outcomes can be across that tremendous scale and volume that we now have.

32. And alongside scale, there is also a separate but equally important development, which is increasing complexity.

33. IP today is no longer neatly contained within single innovations, single inventions or even single brands. It cuts across different domains. One product will have innovations from different companies coming from different jurisdictions, which might then be governed by different frameworks. In trade marks, brand identity is no longer confined to names and logos. It extends into digital environments. The entire ecosystem - not just what you use, but how you use it in conjunction with something else, and increasingly dynamic and AI-generated brand expressions. In patents, a single product may involve multiple layers of innovation - across software, across hardware, data systems, as well as algorithms, often spanning multiple different jurisdictions.

34. In other words, IP is becoming and will become a lot more interconnected, more overlapping, and therefore more difficult to separate cleanly across different categories.

What AI changes, and what it does not

35. This is where AI becomes transformative. It helps us to manage scale and complexity by surfacing patterns, clustering information, and identifying conceptual similarity that might otherwise be difficult to detect manually.

36. At the same time, while AI can help speed up the process, digest information, surface similarities, and give you a good handle on the wealth and breadth of information, it cannot resolve, however, the core question of relevance. That still remains very much a work of art. AI can show proximity, but it cannot quite determine significance yet.

37. Whether in trade marks or patents, the fundamental question remains deeply human: Is this mark distinctive, or is it going to be confusing to the next user, the next customer? Is this innovation truly innovative and novel? What is novel about it? What are its features? How do we determine that? I think it is very much a question of art, not just science. How will consumers, from their perspective and their lens, understand it? How will fundamentally a judge, sitting in the court, should there be a case brought for breach, interpret this? How do you digest the wealth of authorities across different jurisdictions as well as the distinguishing factors, and apply that to the case here?

38. So, I believe that these are not questions that can be answered by searching alone or by AI - at least for now. They require judgment, intuition, maybe a cultural context even, experience, and a sheer human eye that looks out, through experience, for that distinguishing factor.

Opportunities in ASEAN

39. We see this interaction of scale, which I talked about, as well as complexity, playing out very clearly in the region - in ASEAN.

40. The ASEAN market remains one of the fastest-growing innovation and consumer markets globally. Collectively, today ASEAN comprises over 600 million people, with very rapidly digitising, evolving, and substantially improving economies, as well as an expanding base of startups and technology companies.

41. As businesses scale across ASEAN, this complexity arises not just from the volume and interactions between different types of IP, but also from the fact that the region operates across different legal systems - some are civil law, some are common law. There will be differing examination practices - their SOPs will be different, and of course, the enforcement framework, the powers, and maybe even the ability as well as the conviction to pursue enforcement, could well be different.

42. It means that IP work is increasingly cross-jurisdictional in nature. A single commercial strategy may well involve multiple moving parts - coordinating filings across different jurisdictions, portfolio management, and enforcement, across multiple different legal systems, different frameworks that do not always align in procedure or even in interpretation.

43. So, lest you think that I am painting a picture of ASEAN being such a complicated, convoluted, unequal place, actually I am not. What I am suggesting is that with its growth, with its trajectory, and the fact that the ASEAN economy is really growing at a pace that is perhaps one of the fastest in the world, I think this is where opportunities can arise.

44. For practitioners, there is a growing demand for professionals who can seize the opportunity to practise and operate across these differences - managing regional portfolios, and translating across legal systems into a coherent ASEAN-wide IP strategy.

45. For IP offices, it reinforces the importance of cooperation - from examiner exchanges and shared learning to greater interoperability in search, in classification, as well as in examination practices. Initiatives, such as the ASEAN Patent Examination Co-operation (ASPEC), already reflect this trajectory.

What this means for practitioners and examiners

46. More broadly, what this all points to is a shift in the value proposition for all of you as IP professionals - whether in IPOS or operating in the private practice space.

47. As AI takes on more of the basic work in search, in the classification, and doing the summarisation work, the human role moves up the value chain. It moves up the value chain because it allows you to now, through the help of AI, digest more and apply the human experience, judgment and intuition to the gallery of information that AI generates for you. Your role moves both upstream as well as downstream. Upstream, in framing the right questions - with every AI tool you need the right prompt, identifying the real issues, and structuring the analysis that we get. But also downstream, in exercising judgment, anticipating disputes, and translating legal outcomes into commercial strategy.

48. The professionals who are able to thrive in this environment and synthesise and integrate all of these different aspects across an environment that is vast, disparate and uneven, but growing, will not just be those who can do the search better. They will be those who can interpret these results better, understand the concept behind it, synthesise the strategies that will make it work for the consumer and advise the clients better. And more importantly, those who understand when AI is sharpening insight and when it is narrowing perspective, to be able to discern even the outcomes that AI generates for us. I think that is key.

Conclusion

49. So, I hope you have a good discussion later this afternoon with these topics. I framed it in a way which I hope encourages you to think about the broader ASEAN region as a place from which Singapore can serve - operating in Singapore, or from Singapore, or through Singapore - and I encourage you and your clients to think about that.

50. So let me close. You all know that this year's World IP Day theme is IP and sports, and that World Cup is starting tomorrow. So let me use a football analogy.

51. Football today is faster, more data-driven, and more analytically sophisticated than ever before. Some, like me, might say it takes the fun out of watching the sport. Nobody wants to watch another VAR (Video Assistant Referee), digest the goal, draw lines, and decide whether a little bit of your fingernail is offside and so on.

52. But that is the reality. Players and teams now have access to real-time analytics. There are AI-assisted insights and advanced performance modelling. Everyone at the World Cup tomorrow will be wearing all these wearables that are full of technology. But at the same time, the matches that you will watch in the course of the next 5 or 6 weeks will still be decided in moments that cannot be fully predicted or indeed automated. There will be a flash of brilliance, there will be a moment of madness, and there will be these human emotional moments, and they decide and define the game.

53. That is what separates good teams from the great ones, which is making that judgment under pressure, having that instinct, having that intuition, and having come from experience, being able to manifest that on the pitch. Technology can sharpen preparation. It can improve positioning. It can enhance your decision-making. But it does not score the goal. People do that.

54. I think that is the same analogy as I would apply to the industry that you are in. It is fast moving, it is technology-laden, you are going to have more and more different pieces of technology applying for you. But at the end of the day, you sit in the middle of the crossroads of all these technology pieces, and you make that judgement as the human in the middle.

55. Therefore, in that same way, AI will change the way you work and your work as well, and will continue to reshape IP work - making it faster, more data-rich, more complex. But the credibility of the system lies with all of you - to exercise judgment with consistency, with discipline, and with integrity.

56. We are, therefore, in that sense, all on the same team. Whether we are working in the Government at IPOS or MinLaw, out in practice, examiners, practitioners, the whole IP community, what we want to do in Singapore, and for Singapore, which we hope to encourage, is to fundamentally encourage a trusted, high-quality, forward-looking IP system that can support innovation and enterprise. I want to emphasise the word "trusted", because today with technology and AI, you will find that there will be many other jurisdictions that will be able to level up using technology, using AI, using even generative and agentic AI.

57. But what is Singapore's true value proposition? It is that it is a jurisdiction, and it is a profession, that people can trust. That is the key ingredient that we need to always preserve. We need to ensure that - and I go back to my football analogy - even as the game is evolving, as the pace is increasing, as we continue to have more and more resources of a technology nature, that we do not ignore the one element that has made Singapore thrive -that is trust in the system. Where we can have users' trust in the system, we will do well, we will be able to harness all that technology for a positive end outcome.

58. So I encourage you to use an occasion like this, because trust is not built behind a screen, it is not built behind digital communications, but really built in face-to-face interactions, like today. So, continue to openly engage, exchange perspectives, share ideas, exchange best practices, and by doing so, we can ensure that Singapore's IP regime as a whole and our entire profession collectively remains relevant, effective, globally respected, and trusted.

59. Before I close, I want to thank all of you for being here, for being part of this dialogue, and for the important work that you do. You know that IP is one of our hubs that we have been building in Singapore. A small country, but with a far bigger and deeper global presence. I hope that you find today's discussions useful, and the networking even more useful, and I hope to interact with you over coffee in a short while.

60. Thank you very much.

Last updated on 10 June 2026