Transcript of Minister for Foreign Affairs Dr Vivian Balakrishnan's Engagement at Asia Society, New York City, 25 September 2026

Asia Society Global President and CEO Kevin Rudd: Vivian, welcome back to Asia Society. You are one of our oldest continuing guests.

Minister Vivian Balakrishnan: With the most grey hair.

Rudd: No, mine is white. That is the stage you go to once you have been through grief.

Minister: Now you have got two old men.

Rudd: You have now been Foreign Minister of Singapore for 12 years.

Minister: 11 years.

Rudd: So that must make you the veteran Foreign Minister in the region. So, my core question, hard policy question to start with, is: Are you having fun, mate?

Minister: The short answer is yes, for a couple of reasons. First, this is an incredibly dramatic time to be alive. For those of us either on the analytic side of the house, or on the policy and operational side of the house, it is an incredible time of change, disruption, a revolution in technology, and particularly from an Asian perspective. Just reflecting, this society was founded in 1956. At that time, China’s share of (global) GDP must have been about 1 or 2 percent, India in that same league. We live in a completely and dramatically changed world. So, from an intellectual point of view, it is like having a ringside seat to a once in a lifetime, certainly from our perspective, set of tectonic changes. So, no, I am not bored. I am not short of challenges, or near-death experiences. But seriously, intellectually, policy-wise, operationally, and politically, this is a time of exhilarating change.

Rudd: The changes certainly are of an order of magnitude, as I try to conceptualise it. The massive change unfolding in geopolitics, China, the United States. China, moving from nowhere much to centre-stage against all power matrices, the massive disruption represented by new technologies, of which the kingpin is artificial intelligence (AI), extending now across all domains of society, the economy, and state power as well. Thirdly, the one we do not see unfold every day until we have dramatic incidents like we had most recently in Nepal. (Former US Special Envoy for Climate Change) Todd Stern is here with us, in terms of his extraordinary contribution to the climate change debate and the Paris Agreement in 2015. Climate change rolls on relentlessly, and we are in the middle of these three things, and as the movie said, “Everything, everywhere, happening all at once.”

Minister: Do not forget, we also just emerged from a global pandemic.

Rudd: And we have just come out the other side of what could have been the Black Death, except we did manage to pull it out of the fire. But it is not just a ringside seat, my friend, you are also “in the arena,” to paraphrase Teddy Roosevelt.

Minister: And well, so were you.

Rudd: Oh, I have been in and out, and doing a few things.

Minister: He was only the Prime Minister (of Australia).

Rudd: There is a symmetry to our careers. Starting with the economy and looking at the extraordinary dynamism of the region and what is happening in Southeast Asia- give us your sense of growth opportunities and where things stand, not just for Singapore but for wider Southeast Asia, because it is a remarkably strong narrative in the midst of some global economic uncertainty.

Minister: Well, I would characterise it as a counter-intuitive phenomenon. I would ask all of you with your phones, go and check on the first half of 2026, the growth rates for ASEAN countries. I encourage you to do so because we live in a day and age where you must double and triple check everything. You will find that top of the league is Vietnam growing at 8 percent. And then you have got Singapore, Malaysia, Indonesia, growing at about 5 percent. In the case of Singapore, for us to be growing at 5, or even 5 plus percent, given our already high per capita GDP, is enough to raise eyebrows. And mind you, all this is happening in the midst of the initial Liberation Day tariffs announced last year, and now (Section) 301 investigations. Basically, a world now with tariff and non-tariff barriers, a world with disrupted supply chains, a world with wars, and essentially an almost complete closure of the Strait of Hormuz and Bab al-Mandeb. So, for Southeast Asia to be generating that sort of figures, at this point in time, you have really got to scratch your head and ask what is going on.

And my tentative explanation is that, clearly, the revolution which you spoke about – AI, semiconductor, digital. But it is not just that. It is also the biotech and energy revolutions. For those of us who have made it into that competitive league, or at least are part of that global supply chain, there remain major opportunities in the midst of major challenges. So that is my interpretation of the headline figures as of now.

Now, can things change? Of course, it can change. Then you have to ask yourself what are the factors that could change. And one big factor right now is what is happening in Washington, because the push and shove between the two premier superpowers can make, will make, a profound difference to us.

And coming back to Asia, if you take a place like Singapore. Our per capita GDP rose from 500 US dollars in 1965 to now well over 80,000 US dollars per capita in 61 years- quite a dramatic change. Of course, we will give full credit to hardworking and disciplined people and a competent, honest government. But there are some external factors that turbocharged this growth: Pax Americana, an America that won the Second World War, was 40 percent of global GDP, and in an act of enlightened magnanimity, rebuilt the countries it vanquished- Germany and Japan, made access to technology and its market available, the growth of multinational corporations, and global supply chains. If you look at a place like Singapore, we evolved exquisitely to take advantage of it. So did Japan. So too in the earlier years, Hong Kong, Taiwan, Korea.

But the ultimate winner of that Pax Americana and global supply chains and investment-driven, export-oriented economy was the peer superpower called China. The second turbocharge for Singapore is 1978 with the reform and opening of China, which obviously turbocharged China’s growth itself, but it also made a big difference to us. So the philosophical challenge for us is that having evolved exquisitely well to take advantage of the best of both worlds, what is happening in Washington now with that tête-à-tête, whether they get along or whether they fracture, makes a profound difference to all of us. Hence, we will count our blessings, but we cannot afford to take things for granted. That is the mood in Asia.

Rudd: Let us in our conversation go on to artificial intelligence, and come back to geopolitics and the events in Washington later, because I am fascinated by how you in Singapore, as a sophisticated regional economy, sophisticated global economy, are approaching the AI opportunity, and how, in aggregate, you are approaching AI risk, because all of our political systems are grappling simultaneously with both. Give us some insights, Vivian, as to how Singapore is wrapping its head around this.

Minister: Well, let me first, if you do not mind, take a digression at a technological level. Underlying all this fancy AI, deep learning, machine intelligence is basically a mathematical technique: partial differentiation, backpropagation of errors. But basically, what has happened is that we now have a wonderful system for pattern recognition. That is what neural networks enable in human beings, in biology, and now in machines. So just remember, it is pattern recognition, and pattern recognition is also the basis for language. It is also the basis, to some extent, for reasoning, and therefore for constructing a fanciful chain of thoughts, logical analysis, and so on and so forth. Now, that is at the mathematical level.

At the global level, what is interesting is that the mathematical and computational basis for AI, the foundations that both America, China, and the rest of us are operating on is essentially the same stack. The mathematicians, the techniques, the machines, in fact, are the same. This is very different from the Sputnik moment of the contest between the United States and the Soviet Union, when in fact there were, to a significant extent, two different technological and economic systems, ecosystems. Today, in fact, we are all working on the same stack. And because it is the common stack and it has been open and competitive,  in fact, the rate of acceleration has been dramatic. Now, the next point is where the politics comes into, because clearly, if you have something as profound as electricity, you have all the ingredients for another industrial revolution, and the political aspect about understanding industrial revolutions is that those who get it first will have enormous economic advantages, and it is a very short hop from economic to military advantages, and also then having strategic interests to protect. So we are in such a moment.

If you were to try to construct a league table, I think right now America has the lead. In terms of frontier models, maybe six months. If you look at biotechnology, I think it is really neck and neck. If you look at renewable energy, China is way ahead. These are the foundational technological step changes which are occurring, and therefore we should not be surprised, as we were informed by the last industrial revolution. And you know, for people to be trillionaires now, we should not be surprised because this is the start of the Gilded Age of a new technological revolution. But again, if you go back to history, it will have economic and social impact. And you know, since both of us have been in political systems and having to win elections, in the last industrial revolution, the compensation for the Gilded Age was a welfare state, socialism, communism, and before the wars, also extreme right-wing politics. So can you see the plot is being recycled, and we should not be surprised by all that.

Rudd: I found your articulation of the underlying mathematics, pattern recognition, where it then goes to the universality of the stack that both sides are dealing with, and therefore the consequences which flow, and therefore the narrowness of the gap, fascinating. It is a great presentation, and then you are building on that to say, and therefore the Gilded Age. This is a Rockefeller institution. No one was more gilded than Rockefeller. And then they worked out you need to be philanthropic as well. Hence the Asia Society, hence the Rockefeller Foundation, hence a thousand other things in this city.

Minister: Yes. But can you see history rhyming?

Rudd: That is true. So where does history rhyme into the future on these grand, shall I say, social and political themes?

Minister: Well, I was going to look at it from an Asian perspective, and the reason why we in Asia are generally speaking English is because Asia missed the last industrial revolution. In fact, the first part of Asia to embark belatedly in the Industrial Revolution was Japan, the Meiji Restoration. Because China missed it, China underwent a century plus of humiliation and unequal treaties. This time, China has absolutely no intention of missing it again, and because China is at the edge, the frontier edge. You need to understand it is not just a matter of competition with America, being number one or number two, and I am projecting into China’s motivations, but China says, “This is my turn, my time in the sunshine. Why should anyone get in the way, especially when I have the human talent, the brain power?” So can you understand from China’s perspective, I am not going to make the strategic mistake made a century and a half ago. This is a strategic opportunity, and I do not want anyone to get in my way. And all I am asking for is a level playing field, and access to technology, and to get ahead as quickly as I can.

For America, the point is, in fact, it has only been 250 years, and America has had enormous benefits in those 250 years. America has resources and technology, and just as important, the business ecosystem that quickly converts technology into products and services and companies and a profit generating machine. So, this is really a momentous time. But the point is that China is not as different as it is made out to be from America, technologically, mathematically, and even the business ecosystem. Now, of course, there is a difference between the Chinese Communist Party (CCP) rule of China and the democracy which you have here. But can you see that the underlying building blocks of competitiveness are actually the same?

Rudd: Certain motivational structures you attribute to the CCP, I think are well expressed. We often see it through the primary lens of the contemporary strategic competition against the United States, with the object of who becomes numero uno, and that is alive and well in Beijing, and it is alive and well in Washington. And as realists, you cannot wish that away.

Minister: I would inject another perspective here. What does a Chinese entrepreneur worry about the moment he wakes up? Is he or she thinking about competition from America?

Rudd: No, it is domestic competition.

Minister: And if you have been to China, the real spur to competition is domestic. It is cutthroat.

Rudd: That is true, but he also worries about whether an anti-corruption campaign is going to come and bite him.

Minister: Well, that is another phenomenon.

Rudd: That is something we can touch on.

Minister: But the point of view from technology and business, it is the domestic competition in China which is making them so hyper competitive, and then you have got the economic phenomenon. If you ask yourself, what is America’s share of GDP that consumption makes, and you compare that with China. Here is where you see a divergence.

Rudd: That is true.

Minister: Right? The share of consumption as a percentage of GDP in China is much lower than the share of consumption in America. Now here is where you get another apparent contradiction because you have a hyper-competitive economy used to exporting goods, and now also increasingly will have to export services, because what are you going to do with all this capacity? But my point is the capacity is built on a good and sound foundation, and they have a right to be competitive. The only question is, there is a fundamental imbalance in the global economy. There was, before AI, when China became the manufacturing centre of the world, and America benefited from low interest rates because the Chinese were recycling trade surpluses into T-bills, and the American consumer at Walmart also benefited from cheap Chinese goods. But we all know that these fundamental imbalances are not sustainable, and we are now at the point where the tides are changing.

Rudd: Do you think President Xi Jinping has reached this conclusion? Because if we look at the unbalanced or imbalanced nature of China’s economic model, (US Treasury Secretary) Scott Bessent has spoken about this a lot. Absent the private domestic consumption proportion of final demand increasing rapidly, then we are going to have the Chinese system continue to rely upon one or two pistons only in its growth model. Most economies at China’s rate of development have rapidly growing private domestic consumption as a proportion of GDP, rapid rise in private fixed capital investment, rapid rise in private residential consumption, all three of which are in the doldrums in China and have been for a while.

What is working in China is therefore state-directed or state-driven investment in its techno-industrial strategy, plus net exports, and this becomes unsustainable. We have just been talking with our European friends in terms of the rolling carnage in German manufacturing today, and going back to one of your earlier frames, not because of a level playing field, but because if you look at the OECD analysis, you have got an average subsidy level in Chinese industrial model of about 20 to 27 percent. So therefore, cutting a sway through European manufacturing, unlevel playing fields plus the underpinning industrial model of China.

Minister: I would push back a bit on that.

Rudd: That is why I said it so directly.

Minister: Yes. Let us assume all subsidies direct or implicit are removed. Do you think that would automatically remove the academic, scientific, technological edge of China? I do not believe so. Supposing, do you think that the hyper competition domestically is going to disappear? It is not. So the fact that goods and services from China keep coming down in price. The easy answer is (to) just blame it on subsidy, but I am saying there is something more going on than that, and we better understand what that source of competitive advantage is, and it is real. Academic, scientific, technological, business competition domestically. Now, having said that, you have also made the point that in most other economic histories, that achievement is usually followed by increased consumption and a rise of the middle class which consumes more, but you realise we are also lulled into that because of what happened in the first 40 years post-World War II, the GI Bill in the United States, which made education diffuse. The fact that there was a rising middle class which was able to consume, it is the Ford Model T factory and the guys who are on the production line who could afford to buy their products.

Rudd: Like a BYD today.

Minister: But there is another tectonic shift which we have not dealt with yet, and that is demographics. China is, in fact the whole of Asia is, well on the way to becoming a super aged society. Now, what does a super aged society, where say a quarter of us are above the age of 65, mean for consumption, for investment, for business models, for technological advancement? So, you know, we can stand here and say, “Well, these imbalances are not sustainable.” But you cannot just wave your hand and say, “Therefore, thou shall consume more.” It is not so simple, and that is why we are also in uncharted territory as far as demographics are concerned, because the world as a whole, not just China and Asia, the world as a whole, has never had to contend with such low fertility, and yet prolonged life expectancy. So, super aged societies: how you save, how you consume, how you invest. We are already in uncharted territories.

The point I am making, therefore, is do not just be loud and just blame China, and Chinese subsidies for the hypercompetition. They are a real technological power and a competitive economic power, and Europe, as an example, has to come to terms with it. Why is the automotive industry in Japan and Europe suffering? The answer is they missed the electric vehicle resolution. Europe in particular, you go back in time a decade and a half. Do you remember the time when they were lulled that diesel engines were efficient and clean, and then it turned out the data was fudged? Japan kept the internal combustion engine- and not completely incorrectly- hybrids are the way to go. China just leapfrogged all that, and today, I drive a Tesla, but my Tesla is made in Shanghai, and it works perfectly well. Then look at the Chinese export figures and their trade surplus in the last one year alone. It has gone up. It has gone down against America, but against the rest of the world, it has gone up. So you know, just follow the money, look at the numbers.

Rudd: Let me push back against your pushback.

Minister: Yes, that is why it is fun.

Rudd: The reality is a complex thing. What I admire about Singapore’s traditional analysis of reality is its analytical robustness. So yes, the objective reality in terms of the size, sophistication, and density of China’s scientific research and development base has advanced as a consequence of conscious Chinese government strategy for half a century. And anyone who looks at it, must applaud it in terms of its material advance. Simply look at what is occurring in any subcategory of scientific publications today, and the number of postings from Chinese scholars, academics, researchers and usually open source. So therefore, the underlying thesis that China’s technological advances has been exclusively the result of industrial espionage is not sustainable against the reality. Does technological espionage exist? Yes, and that is what happens in the real world.

But the second point is, having conceded that, the OECD data is still empirical, and it presents an uncomfortable picture about the level of subsidy. And so, therefore, when you are looking at every country’s economy constantly needing to adjust and adapt, and China’s brilliance at capturing the EV curve in the way in which they did, it is not purely on the basis of their scientific brilliance, their R&D know-how, the ability to leverage an enormously robust domestic market with maximum domestic competition. They are applying through conscious industrial strategy, subsidies, which underpin their model, which is producing political blowback here, hence the tariff war from the United States and the reduction in the trade surplus of America, and on the cusp of bringing about a similar reaction in Europe. So I suppose my pushback against your pushback is these two realities are occurring at once.

Minister: Let me come back to Todd Stern, because we were co-conspirators in the Paris Agreement. And Todd, you will remember the endless sessions we had with the Chinese negotiators and a very fine gentleman, (former PRC Special Envoy for Climate Change) Xie Zhenhua. Even back then, and I am talking about 2011, it was clear that they were on two strategic tracks. On one hand, not to have China saddled with what they viewed as unfair blame and an unfair burden to deal with climate change, which was due to the Industrial Revolution which did not begin in China. But it was also clear back then, and Todd I think you would agree with me, they had also already figured out that renewable energy is a source of economic competitiveness and strategic resilience, and they were pursuing these twin strategic objectives. Whilst in America, frankly, you were still arguing about whether climate change is real, whether it is a hoax, whether the data, the empirical data, is accurate. My point is, sometimes you can end up debating a complete side street, which is irrelevant to the source of competitive advantage.

That is why I am saying. We can argue about subsidies, we can argue about industrial policy, but actually, the real political and economic question, is: regardless of what system you have, does your system generate good governance? The technological question is: regardless of the education system you have, are your people, and especially your young people, equipped with the intellectual wherewithal to master the emerging technologies and reap economic benefits? Therefore, at a global level, we know eventually everything needs to balance. But the complication now is that we have a geostrategic contest, which sometimes is viewed as a zero-sum game. You have got an industrial revolution, and you have got a demographic bust-up. Any country or system that can deal with these three things simultaneously will win.

If you ask me who, I do not know yet. The game is still on, but my main point today is, do not get distracted with easy answers and apportionment of blame. In America, if you had a system which generated good governance, and if you could get a grip on your deficits, and in fact, the deficits have quite a lot to do with industrial policy, trade surpluses, and competitiveness as well. And then making sure that America retains its advantage, and as I said, the key advantage that America had, was that it was the shortest route to converting scientific discovery into technology, and technology into goods and services which made money. And in American economic history, just when you thought it was the darkest moment, they will always end up putting a rabbit out of the hat, riding on a technological wave. That is why, at least from Singapore’s perspective, we do not count America out. It is still capable of pulling a rabbit out of the hat. You guys really love to live at the edge.

Rudd: That is true. That is America. Welcome to the land of the free and the home of the brave, and it often takes you with white-knuckle moments to the edge.

Minister: Yes, absolutely. But for Europe, see, I tell my European friends, if you add up the nominal GDP of the EU-27, plus or minus the UK. In fact, that aggregate number puts you in the same ballpark, if not even slightly higher than China. In fact, there should be three poles, not a bipolar race. But think about it. Why is Europe not apparently as competitive as the United States and China? I tell my European friends, because you are not running an integrated continental-sized economy. America does, China does, and here Europe has the potential, but for a variety of historical and political reasons, you cannot act as a single continental-sized economy with appropriate industrial policy, a unified fiscal and monetary policy. It is enormously difficult for them to compete.

Rudd: The Europeans are still in search of an answer to the great Henry Kissinger.

Minister: They have woken up, and that is good.

Rudd: They have.

Minister: So do not count Europe out either.

Rudd: They are still in search of an answer to the Henry Kissinger question: “Where is the telephone number that you call?” And the European enterprise, given where they were in history, is not a bad one. The ASEAN enterprise is not a bad one.

Minister: No, but you do not forget, the Industrial Revolution began in Europe. That is why we speak English. But the ultimate heir…

Rudd: I thought it was Shakespeare.

Minister: Shakespeare alone would not account for all of us speaking English. It is the Industrial Revolution. But the ultimate heir of the Industrial Revolution is the United States of America. They took it to the zenith, and now you have for the first time in American history a peer power with the wherewithal and the determination to go full steam ahead, and both America and Europe have to come to terms (with it). 

Coming back to Asia and Southeast Asia. What is it that we want? We want access to technology because we learned the hard way that if you are not at the table, you are on the menu. That is why we all got colonised in the last industrial revolution. So, access to technology. Number two, we want to keep supply chains open. Despite all these tariffs and non-tariff barriers, it still makes sense to run level playing fields, fair and free trade systems, and to be vital parts of efficient global supply chains. Third, to reorganise and restructure ourselves domestically so that our people can master these technologies, not to react in a Luddite fashion, because we know where that ends, but master the technologies and to eke out whatever competitive advantage that we can.

Now, speaking again from a Southeast Asian perspective, we do not believe we can compete with America and China, say for developing frontier models. You know, the special sauce of Silicon Valley, New York, Beijing or Shenzhen. Frankly, I think it exceeds that of what Southeast Asia can generate. But we can make sure we are at the edge of maximal deployment of these tools. Meaning delivery of public services, the level of analysis within government, the accountability, the efficiency of government. Can we make sure that every single civil servant, every single department has maximised the use of these tools to produce public goods, to run a fair and competitive domestic system? It means revolutionising your education systems. And increasingly, you realise education does not stop at your first job, and all of us, you and me too, we will have to have a portfolio of appointments and assignments in our lifetime, and each time will require a reboot of the education and the tools and techniques. So the key word for the rest of us, especially in Southeast Asia, is deployment, and we can make a good run of it, and that is what we are trying to do in Singapore.

Rudd: You have got three quarters of a billion people, almost, in Southeast Asia. This actually is a significant deployment and a significant diffusion. On AI and China, you have put front and centre China’s demographic challenge. 1.5 total fertility rate now.

Minister: I think it is lower than that. Anyway, it is low.

Rudd: It is low and heading south. The question I had in my mind, and I do not know the answer to it, is how the Chinese leadership look at AI in this sense. (A), you have a significant youth unemployment problem in China before the full deployment and diffusion of AI across the economy. Where does that go? And (B), at the same time, and perhaps on a slightly longer time frame, we have China’s date with demographic destiny, and AI therefore becoming part of the solution to the demographic crisis which China anticipates. How do you understand this, and how do you see it? Because I know you guys will have a thoughtful view of it.

Minister: Where in the world do you think is the highest density of robot deployment? And if you have been to China recently, you know that is the highest density of robot (deployment). Why is that so? And it is not just because China is short of labour, because as you said, youth unemployment is a real problem. But it is so because that is where competitive advantage is to be had. So China’s focus on AI and, in particular, embodied AI, which is what robotics is about, because that is where your production lines, your most efficient production lines of the future are going to be found. They have to go down that (path) full speed, with or without the demographic bust. And my point is, the key is how you educate, re-educate, and reskill and retrain your workforce, and that is what we are obsessed with in Singapore. Whether you like it or not, you have to restructure your economy, and you have to figure out what to do for your workers. In Singapore, we adopt the position that a job can be described as a menu of tasks. Well, I think each subtask is going to be completely transformed by AI. Does that mean the job will disappear? May or may not, but it is certainly going to be transformed. But our focus in Singapore is the interest of workers rather than just trying to preserve jobs, which you know are going to be disrupted anyway. So, getting that mental clarity- you are not trying to ossify jobs, but you are trying, but you must, at a political and moral level, be focused on the welfare of each worker. The work that the worker embarks on is going to change inevitably, and you might as well try to get ahead of the curve rather than be caught behind the curve, and then wonder: Why is my unemployment going up? Why is my productivity falling? And therefore why are wages falling as well?

Rudd: How do you quantify, in Singapore’s perspective, the nature of the AI risk to security?

Minister: Oh, very high.

Rudd: You have seen (Anthropic CEO) Dario (Amodei)’s article most recently, and leaving aside state-based actors deploying frontier models; just assuming we have a problem with either rogue states, rogue actors, or most particularly, autonomously acting AI agents unleashed in infrastructure, social, economic, or even military, etc; on your Richter scale of how big the risk is, (I would) really appreciate your view. And secondly, the approach. Dario talks about pacing the frontier, huge pushback against that. But that is one line of option. Calibrating that against, how do you ratchet up your cyber defences, or how do you ratchet up your other, as it were, countermeasures in the event of an incident or a crisis. You will have thought these things through as a government. Give us your wisdom on these ones.

Minister: Well, I am not a doomsdayer, but I think the risks are very high. Let me break down this into buckets. And I am an eye surgeon, I am not a computer scientist, but I would say I have more than a passing familiarity with programming, with running agents, and trying to eke value out of AI systems. So the risks fall into three general buckets. Number one, control of AI agents. And anyone who tells you, even the creator of the AI, that he has got absolute control of the kill switch, is fooling you. If any of you try to run agents, you will understand what I mean. The loss of control is a very, very real risk. Bear in mind that AI systems have been brought up on a diet of all that is available on the internet. Every utterance, and it is not just Shakespeare that AI is brought up on. It is the cesspool of what goes on in the rest of the internet, and more importantly…

Rudd: Including my speeches and yours.

Minister: But think about it. AI is an extension of human beings, from Shakespeare to pornography. Why in the world would we expect AI to be more moral or have greater inhibitory circuits than human beings? We know how human beings behave. We know we are all capable of being saints and devils, and here we are creating a system which is faster than us, wider than us, but brought up on our diet, and we expect it to be a saint. And here we are trying to create autonomous systems, and you think you have absolute control. So loss of control is a very real risk.

The second bucket is that you are now putting weapons of mass destruction in the hands of whoever has access. We talked about (the) pandemic, and now with the benefit of hindsight, you can say it was the coronavirus that led to COVID-19. Yes, it was infectious, but it was not particularly lethal. We basically just had a full-dress rehearsal. It was not the Black Death. It was not even the Spanish Flu. Now, biotech or the genome can be reduced to a language with only four letters. It is even simpler than human language. To cook up, once you (have) mastered the genetic code, edit it, gain of function, optimising it. So be prepared for the day and age when rogue actors have access to bioweapons, to chemical weapons. Chemistry is even simpler than biology. So rogue actors are going to be empowered.

The third category of things you need to think about is the social, political impact of these tools. The relationship to truth, regard for authority. When you take a look at a photo or even a video now, you realise all of us inevitably have reduced our credibility quotient for anything we watch or listen to. Now, that kind of social impact and its impact on politics is very real because no matter how crazy you are today, you can find someone else on the internet, and you will be able to manufacture a video to affirm your craziness.

Rudd: Fact, fiction, fact, and opinion.

Minister: It becomes one big flywheel, and then you wonder why we are all fractious and polarised and divided, and we cannot even agree on facts anymore. Do not even talk about respect for authority. You add up these things: loss of control, abuse by rogue actors, and the causticity of social political discourse, the shadowing of thought and the coarsening of discourse. I think we have got a witch’s brew in front of us. And what we need again is clarity.

So the example I like to give is driving a car. If you want to drive very fast, you need a powerful engine, but you also need good brakes, and you need a highway code. You need to understand what red, amber, and green means when you come to a traffic light. You need to decide: do we drive on the right or do we drive on the left? You need to decide on the insurance policies and how do you distribute and price the risk. So anyone today who comes to me and says, “Oh, leave it alone. Do not regulate.” Well, that is not the way the world works. Or anyone who says, “Trust us.” You know, the internet titans. “Trust us.” No, that is not the way it works either. And there is a reason why moral codes and policies and legislation was invented; so that we could take full use of technology, move quickly, but at the same time take appropriate precautions. I should add the other thing we need is seat belts in cars. So I am not a believer in putting a halt to AI development. I am not even a believer in just putting the brakes on AI. You need to understand we want to drive very fast, and it means a powerful engine. It means good brakes. It means seat belts and a highway code.

And since I am here for the UN, I need to make a plug that the only organisation with universal participation and where we all have a say, big, small, technologically advanced or not, is the UN. So the examples I am going to try to push is: Do we need an ITU, International Telecommunication Union, to set standards? Do we need a UN Convention on AI, the same way we had a UN Convention on the Law of the Sea, UN Convention on Climate Change? And yes, you know, Todd and I and the others in the room struggled through countless hours trying to get consensus, but I think it is more necessary than ever before. So I do not want you all to go out of this room alarmed and afraid. I think we can drive quickly.

Rudd: To conclude then on this, which you have brought to the centre of the discussion, and to reflect on its partner conversations in Washington yesterday. Given we have two AI superpowers at the moment, and they have been meeting yesterday, and unless I have read the outcome statement incorrectly, the landing point on the AI guardrail seems to be, shall we say, thin. Your thoughts on that and how it meshes with the potentiality or the possibility of a sufficiently robust multilateral or plurilateral regulatory environment, on the one hand, given geopolitics. To use your analogy about cars racing, that one is racing and this one is racing, Team America, Team China, and then doing 120 kilometres down the road, the windows get pulled down and saying, “Can we collaborate to make sure we do not have a crash?” How do you actually wrestle that to the ground, and how are you going to solve it, Vivian, and how is the government of Singapore going to fix it?

Minister: That is beyond the paygrade of Singapore, to fix Washington and Beijing. But let me give you some hope. My staff came to the same conclusion as you. The guardrails are thin, but I said no, I am very glad that even President Trump acknowledged that he had a significant discussion on AI. You need to start somewhere, and I am glad that that discussion is going on. Second, and again, I am going to risk overstepping my time limits. If you look at the scientific history of machine learning and AI development in the last decade, because it has only really been around the last decade, decade and a half. The modus operandi was, we are working on the same stack. We make any discoveries, we publish, we make models generally available because open weights models also allow the possibility of retraining those models or making them more specific for your use case. And then competition, not just in terms of money or in terms of power, but academic competition. I publish, I receive due recognition, and in fact, as a researcher, as an individual, my market value goes through the roof because everyone is in search of the next AI genius.

The point I am trying to make is that actually the route that China is taking right now is not a new route. It has actually been the standard modus operandi in which the AI scientific development has progressed over the last 15 years. So it is not novel. It is not made in China. They are just continuing the academic tradition, which led to AI up to this point. Now, the American frontier models are closed source. The analogy is the difference between trademarks and patents. Now, the whole concept of trademarks and patents is an Industrial Revolution concept. Trademarks or secret recipes, say Coca-Cola. To this day, Coca-Cola, it is a trade secret. What are the specific ingredients?

Rudd: They are bad for you, but that is right.

Minister: It is a trade secret. Close models represent that. I trademark something. You cannot use my name in vain, and I am not going to share anything with you. Patents, on the other hand, mean I publish for the greater good, and so that the technological stack can be advanced. But I also get a temporary monopoly, and that is why patents have an expiry date, so that there is sufficient commercial incentive to make advances, but not to the point where it stifles progress. I think in the same way, trademarks versus patents was an answer in the last industrial revolution. We are going to need something equivalent on how as global humanity, (we) will advance AI, deal with the threats of AI, and then allow us to harvest the benefits collectively. Basically, I am trying to end on a positive note.

Rudd: We could continue this conversation for a while about the relative impact of open weights versus proprietary models on the control question you raised before, but that will be part two of our conversation.

Minister: No, we can take a deep dive into that rabbit hole. I mean, it is something which is occupying a lot of my bandwidth.

Rudd: No, that is true.

Minister: And I am just speaking from experience as a dabbler, as a tinkerer, as someone with some input on policy, and hopefully a clear view of what is happening on the global stage on AI.

Rudd: Ladies and gentlemen, the great thing about Vivian Balakrishnan is that I think he is one of the smartest foreign policy professionals, and bigger than that, that we have seen across the world. Please put your hands together.

Minister: Thank you Kevin, you have been very generous.

.    .    .    .    .

MINISTRY OF FOREIGN AFFAIRS

SINGAPORE

26 SEPTEMBER 2026