Building Trusted AI at Scale Cities Startups & Digital Sovereignty – Fireside Chat Moderator- Mariano-Florentino Cuellar

20 Feb 2026 12:00h - 13:00h

Session report

Speakers

Knowledge graph

In-depth analysis

Session at a glance

Summary

This discussion focused on positioning artificial intelligence in the global context, featuring leaders from the International Monetary Fund, World Trade Organization, and Singapore’s government. The panel explored both the tremendous opportunities and significant risks that AI presents for the global economy and international cooperation.

IMF Managing Director Kristalina Georgieva highlighted AI’s potential to boost global growth by 0.8 percentage points, which could help countries like India achieve ambitious development goals. However, she warned of a “tsunami” effect on labor markets, with 40% of jobs globally affected by AI, rising to 60% in advanced economies. She emphasized three major risks: increased inequality between countries, job displacement without adequate support systems, and potential financial stability threats.

WTO Deputy Director General Johanna Hill discussed how AI is reshaping comparative advantage in international trade, potentially favoring countries with strong capital, data, and computing power over labor-intensive economies. She noted that trade growth could increase by 40% by 2040 due to AI, but stressed the importance of investing in skills, regulations, and digital infrastructure.

Singapore’s Minister Josephine Teo explained her country’s strategy of operating as a “trusted node” in the global AI ecosystem, maintaining principled consistency while navigating great power competition. She emphasized that over-regulating AI to address social inequality would be unrealistic, advocating instead for comprehensive social policies including job transition support, healthcare, and education.

All panelists agreed that managing AI’s transformation successfully requires building trust, establishing ethical foundations, and utilizing existing international institutions rather than creating entirely new governance structures. They concluded that while individual countries must manage their own transitions, AI’s challenges are fundamentally global and require international cooperation and learning.

Keypoints

Major Discussion Points:

– AI’s Economic Impact and Global Growth Potential: The discussion emphasized AI’s transformative potential to boost global economic growth by nearly a percentage point (0.8%), with particular benefits for countries like India. However, this growth comes with the risk of widening global inequalities between nations that adopt AI quickly versus those that lag behind.

– Labor Market Disruption and the “AI Tsunami”: A central concern was AI’s massive impact on employment, with 40% of jobs globally expected to be affected (60% in advanced economies). The panelists discussed how AI creates a complex dynamic where some jobs are enhanced, others eliminated, and middle-tier positions are particularly vulnerable, while potentially creating new opportunities in service sectors.

– The Need for Comprehensive Policy Responses Beyond Regulation: The speakers emphasized that managing AI’s societal impact requires a holistic approach including education reform, social protection systems, digital infrastructure investment, and strengthening social solidarity – rather than relying solely on AI regulations to address inequality and disruption.

– International Cooperation and Trust in AI Governance: The discussion highlighted the importance of building ethical foundations for AI, maintaining trust between nations, and leveraging existing international institutions (like the WTO and IMF) rather than creating entirely new global governance structures for AI.

– Democratizing AI Access and Preventing Digital Divides: A key theme was ensuring AI benefits are widely shared globally, with particular attention to helping developing nations access AI technology and preventing a scenario where only advanced economies benefit from AI advancement.

Overall Purpose:

The discussion aimed to examine how artificial intelligence should be positioned within the global economic and political context, focusing on maximizing AI’s benefits while managing its risks through international cooperation and comprehensive policy frameworks.

Overall Tone:

The tone was cautiously optimistic throughout the conversation. The panelists acknowledged AI’s tremendous potential while being realistic about significant challenges. The discussion maintained a collaborative, solution-oriented approach, with speakers building on each other’s points rather than disagreeing. The tone remained consistently diplomatic and forward-looking, emphasizing the need for proactive, coordinated responses rather than reactive measures. There was an underlying sense of urgency about getting AI governance right to avoid repeating past mistakes with globalization.

Speakers

– Announcer: Role not specified, introduced the panel discussion

– Mariano-Florentino Cuéllar: President of the Carnegie Endowment for International Peace, served as moderator for the panel discussion

– Kristalina Georgieva: Managing Director of the International Monetary Fund (IMF), expertise in macroeconomic stability and digital transformation

– Johanna Hill: Deputy Director General of the World Trade Organization (WTO), brings trade perspective to AI technology discussions

– Josephine Teo: Minister of Digital Development and Information for Singapore, expertise in AI governance and public service integration

Additional speakers:

None identified beyond the provided speakers names list.

Full session report

This high-level discussion on positioning artificial intelligence within the global context brought together three distinguished international leaders to examine both the transformative opportunities and significant challenges that AI presents for the global economy and international cooperation. The panel featured IMF Managing Director Kristalina Georgieva, WTO Deputy Director General Johanna Hill, and Singapore’s Minister for Digital Development and Information Josephine Teo, moderated by Carnegie Endowment President Mariano-Florentino Cuéllar.

Moderator Cuéllar opened by highlighting remarkable global progress—life expectancy increasing from 47 years in 1950 to 73 years today—while noting three critical observations: unprecedented global connectivity, increasing fragmentation and competition, and AI’s potential to lead down divergent paths of either shared prosperity or increased inequality.

AI’s Economic Promise and Global Growth Potential

Managing Director Georgieva presented compelling IMF research on AI’s economic potential, showing that AI could boost global growth by 0.8 percentage points, enabling the world economy to grow faster than before COVID-19. For countries like India, this growth trajectory could make ambitious development goals like “Viksit Bharat” achievable. Deputy Director General Hill’s research reinforced this optimism, projecting that trade growth could increase by 40% by 2040 due to AI adoption.

However, both leaders emphasized that benefits are not automatically distributed. Georgieva identified three critical prerequisites for capturing AI’s advantages: robust public digital infrastructure, active removal of barriers to entrepreneurship, and a youthful, innovative population—all of which she noted India possesses.

The Labour Market “Tsunami” and Employment Disruption

The most striking aspect of the discussion was Georgieva’s characterization of AI’s employment impact: “It’s like a tsunami hitting [the labor market] globally.” Her research reveals that 40% of jobs worldwide will be affected by AI, rising to 60% in advanced economies—a counterintuitive finding that challenges assumptions about which economies face the greatest disruption.

The employment effects prove more nuanced than simple job creation or destruction. IMF research in the United States found that while AI creates net positive employment—with one AI job generating 1.3 total jobs—the benefits are unevenly distributed. Georgieva noted that “one in 10 jobs already requires additional skills,” with a smaller segment of workers gaining access to higher-paying opportunities while a larger segment finds employment in lower-paid service sectors. Most problematically, middle-tier workers face the greatest squeeze.

Entry-level positions face particular vulnerability because they tend to be routine and easily automated, potentially disrupting traditional entry points into the labor market.

Reshaping Global Trade and International Cooperation

Deputy Director General Hill provided insights into how AI is altering international competitiveness, with traditional comparative advantage based on labor costs giving way to advantages rooted in capital, data, and computing power. Despite these challenges, Hill expressed optimism about the trading system’s adaptability, noting that trade agreements are “set up so that goods trade and services trade can develop with AI.”

Hill referenced how the WTO’s framework, developed alongside the World Wide Web’s emergence thirty years ago at CERN, provides a foundation for AI-driven trade. However, she acknowledged that some AI areas remain “still too new and still too nuanced” for current systems.

Rather than creating entirely new international institutions, the panelists advocated for adapting existing frameworks. Cuéllar noted that early post-ChatGPT discussions about creating an “International Atomic Energy Agency for AI” have evolved toward more realistic approaches building upon existing capabilities.

Singapore’s Strategy as a “Trusted Node”

Minister Teo offered insights from Singapore’s experience navigating AI development amid great power competition. Singapore’s strategy centers on operating as a “trusted node”—maintaining a “pro-Singapore” approach through consistent, principled decision-making based on national interests rather than alignment with particular powers.

Teo emphasized that maintaining trust requires consistency and principled behavior, noting this is “not a matter of size” but of commitment to reliable governance. She used Singapore’s technology choices as examples of making decisions based on established frameworks while remaining adaptable.

The Centrality of Trust and Comprehensive Policy Responses

All panelists converged on trust’s fundamental importance in AI implementation. Minister Teo argued that success fifteen years hence would be measured by whether citizens trust that AI hasn’t robbed them of livelihoods, safety, or security. She explicitly warned against over-expecting AI regulations to solve fundamental issues like social inequality, advocating instead for strengthening social solidarity through comprehensive policy measures.

Managing Director Georgieva emphasized building strong ethical foundations for AI as a “force for good versus force for evil,” noting that while significant progress has been made on AI’s technical dimensions, much less attention has been paid to establishing appropriate guardrails.

Georgieva outlined three specific policy conclusions: revamping education systems to teach people “how to learn” rather than specific skills, strengthening social protection systems for communities experiencing dramatic changes, and creating enabling environments for AI adoption.

Learning from Globalisation’s Mistakes

A sobering thread throughout the discussion was learning from globalization’s failures. Georgieva explicitly warned against “sugarcoating” AI’s impacts, noting that globalization generated political backlash because its benefits were widely shared while costs were concentrated in particular communities that received insufficient policy attention.

This historical perspective informed the panelists’ emphasis on proactive approaches to managing AI’s transition effects, recognizing that countries adopting AI quickly could perform twice as well as those lagging behind, creating what Georgieva termed an “accordion of opportunities.”

Global Implications and Future Outlook

The discussion highlighted AI’s potential to exacerbate global inequalities while also creating opportunities for developing countries through international cooperation. Hill noted the importance of ensuring AI democratization becomes a priority in development strategies.

Cuéllar emphasized that the challenge is “truly global,” requiring international cooperation rather than individual country solutions. The panelists identified regions like Southeast Asia as potential “laboratories” for innovative AI governance approaches.

Conclusion: Managing Transformation Wisely

The conversation concluded with cautious optimism emphasizing both AI’s transformative potential and the critical importance of wise implementation. The panelists agreed that success depends on building trust, establishing ethical foundations, and utilizing comprehensive policy approaches extending far beyond technical regulation.

Their consensus around leveraging existing international institutions while strengthening social solidarity measures suggests a mature approach to AI governance that balances innovation with social cohesion. Most importantly, their recognition that AI’s challenges require international cooperation and mutual learning provides a framework for addressing this defining technological transition.

As Georgieva noted in closing, the goal is ensuring AI becomes a force for good—a challenge that will require the same international cooperation and commitment to shared prosperity that has characterized successful global policy initiatives. The conversation reinforced that while AI presents unprecedented opportunities, realizing benefits while managing risks demands proactive, comprehensive, and globally coordinated responses.

Session transcript

Announcer

Now we move to a conversation about how artificial intelligence needs to be positioned in the global context. And we have very elite panelists for this session. Ms. Kristalina Georgieva, the Managing Director of the International Monetary Fund. From macroeconomic stability to digital transformation, she’s been a leading voice on how AI will reshape the global economic order and what policymakers must do to ensure that its benefits are widely shared. Ms. Johanna Hill, the Deputy Director General of the World Trade Organization, bringing the trade perspective to a technology that is redrawing the boundaries of comparative advantage. Ms. Josephine Teo, the Minister of Digital Development and Information for Singapore, a nation that has become a global benchmark for how governments can integrate AI into public services.

And this conversation will be held in a few minutes. This will be moderated by Mr. Mariano Florentino Cuellar, President of the Carnegie Endowment for International Peace. So we have a very elite… set of panelists who are going to join us on this panel discussion, which is titled AI Needs to be Positioned in the Global Context. May I please invite our panelists to please join us on stage? So over to you, Mr. Quayar.

Mariano-Florentino Cuéllar:

Thank you very much and good afternoon, everybody. How are we doing AI summits? Let me try that again. Hello, Delhi. Thank you. Much better. It is not every day that we have the pleasure of having such a distinguished panel of international leaders. And I want to start by making three observations only as special observations for those of you who have chosen to be with us this afternoon. You could be anywhere in this complex, anywhere in the city, and you’re right here with us. The first is about the role of technology and science and global ties in making the world better. For those of you who are younger than me, which is most of you in the audience, you will live longer than my generation because of global ties, commerce, science and technology.

In 1950, when India was a young nation, global life expectancy was 47 years. Now it’s closer to 73 years. But at the same time, the second point is that the world that we are navigating today is fragmented. That set of global ties, diffusing science and technology, advancing global understanding and cooperation is a lot harder now than it was even five or 10 years ago. And everybody who’s been on this stage has been alluding to that in some way, that reality. The third point is that the use and development of AI will have an effect on those ties and on that prosperity in all likelihood. But there are divergences, different paths around AI. Some countries are using it more, some less.

Some countries play a certain role. Some very developed role in the tech stack and others less. To talk about these issues, I cannot imagine a better pan. It’s not every day, as I said, that we have the managing director of the IMF, the deputy director general of the World Trade Organization, and the minister for information and digital development from Singapore. So I’m going to start with a question for managing director Gorgieva. And the question is, all this discussion about artificial intelligence at the frontier, what do you see as the greatest possibilities and the greatest risks?

Kristalina Georgieva

Thank you very much. Namaste. Namaste. AI is an incredibly transformative we know. And the question is, what does it do for the world economy? We did some research, and here is the answer. Based on what we know, AI can lift up global growth by all. Almost. a percentage point, we say 0 .8%. What does that mean? It would mean that the world would grow faster than it did before the COVID pandemic. And that is fantastic for creating more opportunities, more jobs. This is the magnitude that we see for India. And it would mean that India’s Vixit Bharat is achievable. It also means that the world risks to be even further diverse. The accordion of opportunities may open even more from countries that do well to those who fall behind.

Thank you very much. Actually, what we see is the potential for countries that go fast on digital infrastructure, on skills, on adoption of AI, that they can do twice as well as those that don’t. So what is our main reason to be here at the AI Summit in Delhi? To embrace India’s proposition of democratizing AI, making sure that experience in India can then be passed to other countries, especially countries in the developing world, to make diffusion, to make adoption of AI. The main priority and do it with focus on people, on improving the opportunities, the livelihoods of people. I am very optimistic about AI. I’m also not naive. It brings significant risks. First, it brings the risk of making countries and the world less fair.

Some have it and others don’t. Second, it brings the risk of displacement of jobs with no good thinking about how to help people find their place in the new AI economy. We calculated this risk as very high. We actually see the impact of AI on the labor market like a tsunami hitting it globally. 40 % of jobs will be affected by AI, some enhanced, others eliminated. Emerging markets, 40%, but in advanced economies, 60%. And that is happening over a relatively short period of time. And the third risk we at the IMF worry a lot about is financial stability risk. Could AI get loose and create havoc on financial markets? But on balance, my appeal to all of us is embrace the opportunities, be mindful of the risks, and manage them well.

And above all, make sure that the spirit here is that AI is for the well -being of everybody, everywhere. Thank you.

Mariano-Florentino Cuéllar:

And what we’re going to do, we’re going to. I’m going to come right back to these questions in a minute, but I want to bring in the Deputy Director General of the World Trade Organization into the conversation. I want to ask you, picking up exactly where Managing Director Gheorgheva was going. the interest in democratizing the technology, having more countries be closer to the frontier. For more than a generation, as you know, we have been having arguments about trade globally and about whether trade helps reduce the gap in well -being between countries or actually pulls them apart even more. And given all that experience, I wonder what role you think the international trading system has in dealing with potential inequities and access to AI and the development of AI.

Johanna Hill

Thank you so much for the invitation. To be here, definitely we see that trade can help the diffusion of AI to those that most need it. And we also think that AI can help trade and can help lower income and middle income economies really progress through trade. Now, we do see that AI is really shifting what we think of as comparative advantage to those economies that are more strong in capital, data, and in computing power. and therefore the countries that are more labor intensive feel more at risk. At the same time, we also see important opportunities for these same countries. Of course, with all the caveats that we’ve been speaking about, the importance of investing in skills and regulations and in infrastructure, digital infrastructure are incredibly important.

Our research suggests that by the year 2040, trade growth could be almost growing by 40%. So we see really important opportunities for the middle and lower income economies. And trade is already working well in that way. Our trade agreements, the world trading system is set up so that goods trade and services trade can develop with AI. But there are some areas where they’re still too new and still too nuanced. And we still have to wait and see how that will develop and how the system has to accommodate.

Mariano-Florentino Cuéllar:

Minister Teo, as that system evolves, and we deal with this, emerging, not even emerging anymore, emerged technology. we talk about how much it’s going to affect countries large and small. You are playing a critical role, and I know you’re playing a critical role because I see you at every single AI summit in the world. It’s amazing. But how are countries like Singapore in a position to navigate this tsunami, these changes? And what, in particular, what do you think we could learn from Singapore’s strategy, as I see it, of being at the forefront here on AI governance, the Model AI Governance Plan, for example, but also navigating a world that some people see as balkanized between China and the United States around the technology stack?

Josephine Teo

Thank you very much, Tino. That’s a lot of questions packed into one. I’ll do my best to address them. I think embedded in what you’re saying is that there is the risk of technology decoupling. And what does a small state do? In this kind of context? And how do we navigate the big power contestation? The way we think about it is that for Singapore, it’s very important for us to maintain this ability to operate as a trusted node. Trusted node means that, well, we can trust you with our technology. So your companies, your people can continue to access this, whatever is the most sophisticated, because they will not be abused and the risk of them being misused is also minimized.

The question, however, is how do we remain trusted? And I think the only way to do so is if we act in a consistent and principled way. And being consistent and principled is not a matter of size. And Singapore is not the only small state that has a good track record of holding this discipline. We are consistent in being. Pro -Singapore. And sometimes our choices may align with this country or that country. Sometimes they will align with many countries. Sometimes they only align with a few countries. But they always align with our own interests. In technology choice, for example, 5G, we are always operating on the basis of principles. Number one, that these are commercial decisions that have to be undertaken by the operators of the mobile networks.

And they have to decide on the basis of what works for them in terms of performance, in terms of security, in terms of resilience, keeping in mind what are all the rules that are in place in our context. So those are the broad directions in which we operate in. And it’s not easy, but it’s a path that has served us well.

Mariano-Florentino Cuéllar:

And I note that among the many things that Singapore, I think, has contributed to the discussion of AI globally, in addition to being a trusted node and connecting different countries, there’s also the role Singapore and the region of Southeast Asia plays in all this because Southeast Asia is such a region of such diversity and importance globally. And I want to come back in a minute to the question of how we might imagine Southeast Asia evolving as almost a laboratory for some of the issues we’re talking about. But first, I want to go back to Kristalina, if I may, and ask you about, it was clear in your earlier remarks that you see enormous possibilities for AI.

But you also acknowledge candidly something that maybe not every speaker has acknowledged, which is along with that opportunity will probably come some disruption. Some real policy difficulties in some countries that are experiencing rapid change. The question then is how we might develop the right strategy so that the productivity gains that the world can experience would actually translate into shared prosperity. What do you think we can do on that score?

Kristalina Georgieva

The first thing we ought to do is… to carefully observe what is actually happening and then project what are the implications for policymakers. At the Fund, we did a very interesting piece of research in the United States assessing how much AI is affecting already the labor market. And we found out that one in 10 jobs already requires additional skills. And for those who have these skills, the job pays better. Now, with money in their pocket, people then go and buy more local services. They go to restaurants, to entertainment. That creates demand for low -skilled jobs. And to our surprise, the total impact on employment in the aggregate is positive. One job with AI, 1 .3 jobs.

1 .3 jobs. in total employment. But what does that mean? It means that a smaller segment of people get higher opportunities. A larger segment, yes, they can have jobs, but jobs that are on the lower end of the pay scale. And the most problematic is the fate of those squeezed in the middle. Their jobs don’t change. In relative terms, they pay less, and some of these jobs disappear. What concerns us the most is that jobs that disappear tend to be entry -level jobs. They are routine, and they are easily automated. So if you are in this place of the labor market that is easily automated, of course that creates a risk. Now, we are going to talk about the risk of the labor market.

We are going to talk about the risk of the labor market. We are going to talk about the risk of the labor market. We are going to talk about the risk of the labor market. We are going to talk about the risk of the labor market. We are going to talk about the risk of the labor market. We are going to talk about the risk of the labor market. once obviously we will continue to work with countries to understand what is happening and then how do we project it for policies for the future i would make three conclusions so far and of course we have to be agile in how we look at ai the first one is education has to be revamped for the for a new world people have to learn to learn not to learn specific skills so much and there has to be second there has to be support for those if they’re a big chunk in a particular local economy and this labor market is changing dramatically there has to be social protection social support so they don’t feel like what happened with the industrial world workers in the united states when their jobs were exported overseas and three it is very important that we look at the overall enabling environment.

Why in some places AI makes it faster and in others it doesn’t. And what we find is not very surprising. Some parts of the economy, some parts of society are naturally better positioned because they have digital infrastructure in place. They are already in the digital world because there is more demand for entrepreneurship. Somebody spoke about it and entrepreneurship is more dominant. And I think it is important for the world to be very attentive to what works, what doesn’t work and not sugarcoat the picture because if we do, we would end up where we ended up with globalization. People revolting against it despite all the benefits it brings because, yes, the world as a whole benefited but some communities were devastated.

and the world did not pay attention to these communities in a timely manner. So that is my conclusion so far. And I know that I am very mindful that we are going to learn much more. At the front, we are trying to see how our country is positioned. Some countries actually have more demand for AI skills than supply. Some countries have more supply of AI skills than demand, and some have neither. So we have to work on multiple fronts, and we have to work based on concrete assessment of conditions in countries and localities in countries. I want to finish with a message to the Indian friends here in the audience. You’re very fortunate that your country invested in public digital infrastructure.

So this country… Condition for AI? Check. You are very fortunate because your country is removing actively barriers to entrepreneurship. And on that count, we say check. And you are super fortunate to have youthful, energetic, innovative population that is embracing AI. So what do we say? Check. So all the very best.

Mariano-Florentino Cuéllar:

This is terrific. Perfect. Minister Teo.

Josephine Teo

Can I agree withthe managing director more, if I may be allowed to chime in? I think sometimes there is a desire, a tendency to want to think of ways of regulating AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need. For example, in making… I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk.

I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. I’m not underestimating the need for AI in order to slow down its advance and perhaps to try and forestall the risk. But to over -expect AI regulations to deliver on the other important issues, such as the potential for greater social inequality, I think it’s unrealistic. The way to deal with it is to look at what other methods there must be to strengthen social solidarity.

For example, what provisions do we put in place to help people to move from one job to the next? What provisions do we put in place to ensure that even people who don’t earn a lot have the prospect of owning their own homes, access to good health care, educating their children to a very high level? I think these are the other things, and you cannot run away from those conversations just by expecting regulations to solve the problem.

Mariano-Florentino Cuéllar:

So what I’m hearing you both say, in a way, is that it would be a very silly thing if we tried to solve health care problems. just by regulating pharmaceuticals. That would be a very poor fit, right? At the same time, you recognize that, you know, certain products that are sold, it’s good for them to be safe. And in fact, safety, trust, security can make them even more easy to diffuse. But I think what a very important takeaway from both of you is that the entire spectrum of tools that a society has to build social cohesion are going to be important in the transition to a more AI -driven economy. And we shouldn’t ignore them, but we shouldn’t put just the focus on what we can do by making models built in a certain way.

And I’d love for you to chime in because trade has come up already, just even in the last like 47 seconds of a bunch of times. Actually, yes.

Johanna Hill

We put out a report last year that looks at this issue exactly in that way. We look at the opportunities that I talked about of AI in the future, not only for the advanced countries, but developing in the lower income. But we also look at the need for national policies for that to actually… happen and to help transition. And so we look at issues around competition policy, around labor force. around skills development, around education. And to do that, the world trading system cannot do it alone. We need to partner at our level with international organizations and at the national level with the appropriate authorities and the private sector in order to have that holistic approach.

I would say lesson learned from past experiences, and we definitely want to apply those lessons to this new one.

Mariano-Florentino Cuéllar:

So we have about four minutes left, and I have a last question for you all. Well, imagine yourselves in the future looking back at the past, maybe 15 years in the future. And at that point, you’re being interviewed on the same stage here in India, and you’re saying it’s been a very good thing to see how well the world has handled its relationship with this emerging technology of AI, and it’s turned out very well because blank. And I want you to mention one thing that you think in particular would have been so critical to make that transition well. You’ve all mentioned a bunch of things, but I’m interested in the main, most important takeaway that you’d like to leave the audience with.

For me, that one word is trust.

Josephine Teo

In 15 years, if we went and asked citizens in all the countries where AI is being deployed widely, do you trust this technology? If their answer is no, then I believe that we must have failed in some way. If they believe that this technology has been implemented in a way that didn’t rob them of a livelihood, that didn’t rob them of, you know, being totally misinformed about the world, didn’t rob them of, you know, being able to carry out their lives in a safe and secure manner, it didn’t destroy families. I think if they can still say that this is a technology that can work reasonably well if you put in place the safeguards, I think we would have come a long way.

Mariano-Florentino Cuéllar:

Deputy Director?

Johanna Hill

An appreciation for what the world trading system can and is delivering. You know, when I think about it, last year it turned 30 years that the WTO was born. And down the road at CERN, the World Wide Web was being created by scientists that wanted to collaborate. And that architecture, which is technology neutral, allowed for those developments of the digital economy to come through. And how much of that architecture can serve us for this new wave? And then concentrate on those areas that are still needing to be worked on by collaboration, by cooperation, and focus on those. You know, trading with trust, trading with safety, and then appreciating and using what we already have to deliver.

Mariano-Florentino Cuéllar:

Managing Director?

Kristalina Georgieva

Well, in 15 years, if my life expectancy has grown by another 50 years, I would say, great, we are successful. But on a serious note, I think, to me, the most important… factor, it goes a bit in the trust area, is the ethical foundation of AI. Whether we would manage to put AI on the foundation of force for good, or we leave space for AI to be force for evil. And that balance is not easy one. When I look at progress so far, we have done much more on the technical side of AI, and much less on building that strong ethical foundation, and putting guardrails that are not restricting innovation, but are protecting us from AI for bad.

I still want my 50 years extra life.

Mariano-Florentino Cuéllar:

One closing observation to just reinforce my appreciation to the three of you and the work we do. So in the weeks immediately after the release of ChatGPT, which seems like 20 years ago, but it was not that long ago, there was talk about the need for an international atomic energy agency for AI or a new international agency or treaty. We don’t talk about that anymore. And I think in some ways it’s an appropriate and mature recognition that we already have a set of institutions and mechanisms in place to deal with a set of emerging challenges. I think it’s also a recognition that many individual countries have to do their part to create social cohesion and manage this change and this transformation effectively.

But I would ask that this audience recognize that all three of our remarkable leaders here on the stage also reflect another reality, which is that even if sovereignty is important and even if individual countries have to have their own priorities, the challenge of how we best live with the technology we have created is truly a global one. It’s not an individual country. It’s a country one. And the conversation we’re having today is an example of how we can learn from each other and find the right solutions. Thank you and namaste.

Kristalina Georgieva

Speech speed

119 words per minute

Speech length

Policy Toolkit for an Inclusive AI Transition

Topics

Social and economic development | Human rights and the ethical dimensions of the information society

Johanna Hill

Speech speed

157 words per minute

Speech length

Policy Toolkit for an Inclusive AI Transition

Topics

The enabling environment for digital development | Artificial intelligence | The digital economy

Josephine Teo

Speech speed

176 words per minute

Speech length

Policy Toolkit for an Inclusive AI Transition

Topics

Social and economic development | Human rights and the ethical dimensions of the information society

Mariano-Florentino Cuéllar

Speech speed

191 words per minute

Speech length

Trust and Global Cooperation as the Bedrock for AI Governance

Topics

Artificial intelligence | The enabling environment for digital development

Announcer

Speech speed

140 words per minute

Speech length

203 words

Speech time

86 seconds

Framing AI in the Global Context

Explanation

The Announcer sets the stage by positioning AI within a global discussion, highlighting its relevance to international policy and cooperation.

Evidence

“Now we move to a conversation about how artificial intelligence needs to be positioned in the global context” [79]. “So we have a very elite… set of panelists who are going to join us on this panel discussion, which is titled AI Needs to be Positioned in the Global Context” [82].

Major discussion point

Opening and Contextualisation

Topics

Artificial intelligence | The digital economy

Agreements

Agreement points

AI will create significant labor market disruption requiring comprehensive policy responses

Speakers

– Kristalina Georgieva
– Josephine Teo
– Johanna Hill

Arguments

AI will affect 40% of jobs globally like a ‘tsunami,’ with 60% impact in advanced economies and displacement of entry-level routine jobs

AI regulation alone cannot solve social inequality issues; comprehensive social solidarity measures are needed including job transition support and access to housing, healthcare, and education

A holistic approach requiring partnership between international organizations, national authorities, and private sector is necessary

Summary

All speakers acknowledge that AI will cause massive labor market changes and agree that addressing this requires comprehensive policy responses beyond just AI regulation, including social protection, job transition support, and multi-stakeholder collaboration

Topics

Artificial intelligence | Capacity development | Social and economic development

Need for multi-stakeholder collaboration and holistic approaches to AI governance

Speakers

– Kristalina Georgieva
– Josephine Teo
– Johanna Hill

Arguments

Social protection and support systems are essential for communities experiencing dramatic labor market changes

AI regulation alone cannot solve social inequality issues; comprehensive social solidarity measures are needed including job transition support and access to housing, healthcare, and education

A holistic approach requiring partnership between international organizations, national authorities, and private sector is necessary

Summary

All speakers emphasize that AI governance cannot be addressed through single approaches or by individual actors alone, but requires comprehensive collaboration across multiple levels including international organizations, national governments, and private sector

Topics

Artificial intelligence | The enabling environment for digital development | Human rights and the ethical dimensions of the information society

AI presents both tremendous opportunities and significant risks that must be carefully managed

Speakers

– Kristalina Georgieva
– Johanna Hill

Arguments

AI can lift global growth by 0.8 percentage points, helping countries like India achieve development goals

Trade growth could increase by 40% by 2040 due to AI, creating opportunities for middle and lower income economies

Summary

Both speakers acknowledge AI’s transformative potential for economic growth while recognizing the need for careful management of associated risks and challenges

Topics

Artificial intelligence | The digital economy | Social and economic development

Trust is fundamental to successful AI implementation

Speakers

– Kristalina Georgieva
– Josephine Teo

Arguments

Building strong ethical foundations for AI as a ‘force for good’ is more important than technical development alone

Success in 15 years depends on citizens trusting that AI technology doesn’t rob them of livelihoods, safety, or security

Summary

Both speakers emphasize that citizen trust and ethical foundations are more critical for AI success than purely technical advancement

Topics

Artificial intelligence | Human rights and the ethical dimensions of the information society | Building confidence and security in the use of ICTs

Artificial intelligence | The enabling environment for digital development | Follow-up and review

Unexpected consensus

Arguments

AI regulation alone cannot solve social inequality issues; comprehensive social solidarity measures are needed including job transition support and access to housing, healthcare, and education

Social protection and support systems are essential for communities experiencing dramatic labor market changes

Summary

While both recognize the need to address AI’s social impacts, Teo explicitly warns against over-relying on AI regulations to solve broader social issues like inequality, advocating instead for comprehensive social solidarity measures. Georgieva focuses more on the need for social protection systems specifically for labor market changes, without the same emphasis on limiting regulatory expectations.

Topics

Artificial intelligence | Human rights and the ethical dimensions of the information society | Social and economic development

Unexpected differences

Role of existing international institutions versus need for new frameworks

Speakers

– Mariano-Florentino Cuéllar
– Johanna Hill

Arguments

The challenge of living with AI technology is truly global, requiring countries to learn from each other rather than creating entirely new international agencies

The existing World Trade Organization architecture, being technology-neutral, can serve the AI era while focusing on areas needing new collaboration

Explanation

While both support leveraging existing institutions, there’s a subtle difference in approach: Cuéllar emphasizes that the conversation has moved away from creating new international AI agencies toward using existing mechanisms, while Hill specifically advocates for the WTO’s continued relevance. This represents different institutional perspectives on how to handle AI governance globally.

Topics

Artificial intelligence | The enabling environment for digital development | Follow-up and review

Overall assessment

Summary

The discussion shows remarkably high consensus among speakers on AI’s transformative potential and major challenges, with disagreements primarily focused on implementation approaches rather than fundamental goals. The main areas of difference involve the balance between regulation and broader social policy, and the specific pathways to achieve shared objectives of inclusive AI benefits.

Disagreement level

Low to moderate disagreement level. The speakers demonstrate strong alignment on core issues (AI’s economic potential, need to address inequality, importance of international cooperation) but differ on emphasis and specific mechanisms. This suggests a mature policy conversation where fundamental principles are established and debate centers on implementation strategies. The implications are positive for AI governance as it indicates potential for coordinated international action despite different institutional perspectives.

Partial agreements

All speakers agree that AI presents significant opportunities for economic growth and development, but they emphasize different pathways: Georgieva focuses on overall economic growth metrics, Hill emphasizes trade-based opportunities, and Teo prioritizes citizen trust and social acceptance as the foundation for realizing benefits.

Speakers

– Kristalina Georgieva
– Johanna Hill
– Josephine Teo

Arguments

AI can lift global growth by 0.8 percentage points, helping countries like India achieve development goals

Trade growth could increase by 40% by 2040 due to AI, creating opportunities for middle and lower income economies

Success in 15 years depends on citizens trusting that AI technology doesn’t rob them of livelihoods, safety, or security

Topics

Artificial intelligence | The digital economy | Social and economic development

Both recognize that AI will create winners and losers among countries, but they frame the challenge differently: Georgieva focuses on the speed of adoption and digital infrastructure readiness, while Hill emphasizes the fundamental shift in what creates competitive advantage from labor to capital/data/computing power.

Speakers

– Kristalina Georgieva
– Johanna Hill

Arguments

AI risks creating greater divergence between countries, with those adopting AI fast doing twice as well as those falling behind

AI shifts comparative advantage toward countries strong in capital, data, and computing power, potentially disadvantaging labor-intensive economies

Topics

Artificial intelligence | Closing all digital divides | The digital economy

Both agree that addressing AI challenges requires comprehensive approaches beyond just technology, but Georgieva emphasizes educational reform specifically, while Hill focuses on institutional collaboration and partnership structures.

Speakers

– Kristalina Georgieva
– Johanna Hill

Arguments

Education must be revamped to teach people ‘how to learn’ rather than specific skills for the AI economy

A holistic approach requiring partnership between international organizations, national authorities, and private sector is necessary

Topics

Artificial intelligence | Capacity development | The enabling environment for digital development

Artificial intelligence | The enabling environment for digital development | Follow-up and review

Takeaways

Key takeaways

AI has significant potential to boost global economic growth by 0.8 percentage points, but risks creating greater inequality between countries that adopt it quickly versus those that fall behind

AI will cause massive labor market disruption affecting 40-60% of jobs globally like a ‘tsunami,’ requiring comprehensive education reform and social protection systems

Success with AI depends fundamentally on building and maintaining public trust in the technology through ethical implementation and safeguards

AI regulation alone cannot solve social inequality issues – comprehensive social solidarity measures including job transition support, housing, healthcare, and education access are essential

Small countries can navigate AI geopolitics by maintaining positions as ‘trusted nodes’ through consistent, principled decision-making based on their own interests

The existing international trade architecture can support AI diffusion, but requires partnership between international organizations, national authorities, and private sector for holistic approaches

Countries with strong digital infrastructure, entrepreneurship culture, and skilled populations (like India) are best positioned to benefit from AI transformation

Building strong ethical foundations for AI as a ‘force for good’ is more critical than technical development alone

Resolutions and action items

Continue IMF research to monitor AI’s actual impact on labor markets and project implications for policymakers

Focus on revamping education systems to teach people ‘how to learn’ rather than specific skills

Develop comprehensive social protection and support systems for communities experiencing dramatic labor market changes

Maintain and strengthen international cooperation through existing institutions rather than creating new AI-specific agencies

Prioritize building ethical foundations and guardrails for AI that protect against misuse without restricting innovation

Unresolved issues

How to specifically prevent AI from becoming a ‘force for evil’ while maintaining innovation

Concrete mechanisms for ensuring AI benefits are widely shared rather than concentrated among a few

Specific policy frameworks for supporting middle-tier workers who are most at risk from AI displacement

How to balance AI advancement with the need to slow down risks through regulation

Detailed strategies for helping developing countries avoid falling further behind in the AI divide

Specific international coordination mechanisms for AI governance beyond existing trade frameworks

Suggested compromises

Balance AI regulation to provide necessary safeguards without over-restricting innovation or expecting regulation alone to solve social inequality

Use existing international trade institutions and frameworks rather than creating entirely new AI governance bodies

Allow countries to make technology choices based on their own commercial and security principles while maintaining international cooperation

This comment draws a powerful historical parallel that adds crucial context to AI policy discussions. It demonstrates learning from past policy failures and warns against repeating the mistakes of globalization by ignoring distributional effects and community-level impacts.

Impact

This historical framing elevated the entire discussion by connecting AI governance to broader lessons about managing technological and economic transitions. It reinforced the importance of proactive social policies and influenced the conversation’s emphasis on trust, social cohesion, and inclusive growth strategies.

Overall assessment

Explanation

She noted that much more progress has been made on the technical side of AI than on building strong ethical foundations and appropriate guardrails

Disclaimer: This is not an official session record. DiploAI generates these resources from audiovisual recordings, and they are presented as-is, including potential errors. Due to logistical challenges, such as discrepancies in audio/video or transcripts, names may be misspelled. We strive for accuracy to the best of our ability.