UN Geneva - Multimedia Newsroom : UNCTAD Press conference: impact of AI on labor, policies and global equity

Good morning everyone.

Welcome in this At this press conference we are going to map the AI and digital economy environment ahead of the the AI for Good summit to be held in in beginning of July 7th to 11th.

Of course, we are part of these events with all the analysis and researchers that has been made in OCTAD.

We are here to remind you on the challenges and opportunities AI represents for developing countries.

And without further ado, I will give the floor to Topion Fredrickson, who is the Head of e-commerce and Digital Economy Branch at ONTAD and who has to remind you about the challenges, opportunities over to you.

Thank you very much, Catherine.

Before going specifically into AII, would like to start by giving a broader perspective on digitalisation, where of which AI is a very important part.

[Other language spoken]

As you know, the digital economy is growing at a very, very high pace, connecting more people at higher broadband speed and offering many new digital services, of which AI is 1.

For example, over the past 20 years or so, we have seen the number of Internet users grow from 1 billion to more than 5 billion.

We have seen the quadrupling of the production of semiconductors in the world.

We are expecting mobile data flows to more than double until 2030 from today and the number of Internet of Things devices will grow from about 16 billion to 39 billion in 2029.

Just to illustrate this strong pace and that we have experienced and is also expected to continue or rather accelerate.

But despite this rapid expansion of digitalisation, there are also very large divides in the world in the area of digital.

For example, while we're in the advanced economies, think about 5G mobile networks as something that basically everyone has.

About 90% of people have access to 5G mobile networks.

Only 3% of the people in least developed countries have access to such networks.

Another illustration is e-commerce.

In the more advanced economies, around 80% of people are shopping online as a as a habit, whereas in least developed countries and landlocked developing countries of the world, less than 10% are generally engaging in such activities.

This has implications for the ability to take advantage of digital opportunities and for countries to use digitalisation for create value, create jobs, to innovate and so on.

So and this is even more accentuated in the context of artificial intelligence.

Of course, another illustration of the current situation in the in the case of the digital economy, you can see a very high level of concentration in the market with especially two countries dominating the scene, the United States and China.

Just to illustrate, if you look at the world's largest companies by market capitalization in 2005, two were in the ICT field.

By 2015, five were in the ICT field and as of now, 8 plus Tesla are among the top ten.

The only one that is not in that field in the tech field is Saudi Aramco, which is of course an oil company from Saudi Arabia.

The companies that are the the leaders in terms of digitalization are also the leaders to harness the opportunities that artificial intelligence brings because the key issues here is having access to vast amounts of data and having access to very strong capabilities in turning those data into intelligence through artificial intelligence.

Another aspect to keep in mind as we go into the global debates about the impact of digitalisation and also on AI is the environmental implications.

We have seen now that greenhouse gas emissions from the ICT sector are comparable to the global aviation industry or the global shipping industry and they are growing fast, very much linked to artificial intelligence uptake, update uptake.

We're also seeing the growing demand for critical minerals, another strong aspect in global debates now as the digital transition, the development of artificial intelligence is relying very much on the same kind of minerals as the green transition.

The the shift towards low carbon technologies and the artificial intelligence is very much driving the increase in the demand for electricity and water.

It is estimated that electricity used by the world's largest data centres will triple between 2018 and but that it tripled between 2018 and 2023 and that global electricity used by data centres will AA yet between now and 2030.

In some places of the world this has a very strong impact on the electricity consumption.

In Ireland for instance, 20% of all the electricity used in the economy goes to the data centres and if you take the capital of Ireland, Dublin, then 50% of all electricity use is by data centres.

We've also seen a rapid growth of waste from digitalisation and it grew by 30% about in the past decade.

And only one quarter of such waste is actually formally collected.

So a lot of this is not handled.

And that means missed opportunities for recovering a lot of valuable components and metals and minerals, but it also means that a lot of hazardous material, hazardous material is not properly taken care of.

A few policy reflections on this.

It's very important to enhance the ability of developing countries to take more advantage of the opportunities that digitalization can offer.

And that means basically to build their digital readiness, as we call it.

And there's also a need to adopt policies at both national, regional and international levels, since digitalization is a very global phenomenon.

Some of the key areas for for global policy making, as we see it, is in the area of data and data flows, competition policies, trade policies, and taxation policies.

We also see a need to foster more circularity in the digital economy.

As of today, very, very few components are possible to repair, to recycle and to refurbish, which leads to a lot of waste.

It's also important to raise the awareness in the private sector, among governments and also among people at large about the environmental implications of the use of digital devices and infrastructure.

The upcoming big events here in Geneva, the the the AI for Good Summit and also the the Wisys Plus 20 review.

They offer opportunities here to have a multi stakeholder dialogue on this, these many issues on how to foster a more inclusive, but also more sustainable digital future.

And here we are stressing a lot that in order to make real changes, effective collaboration is really essential, both across countries, across stakeholders and with a strong involvement of the private sector and civil society.

So I will end there and maybe you'll move through to Arkan.

Thank you very much, Arkan.

You are the head of the I know they're complicated, sorry.

Introduce to.

No problem, Catherine.

My name is Agron Fareth.

I'm the head of the Technology Innovation and Knowledge Development branch in Antar, which is a sister branch of Torbians in the same division on technology and logistics.

And we look at technology from a slightly broader angle.

We look at digital technologies, but also at a broader set of what we call frontier technologies that we think are determining factors for the way the the economy and society is evolving today.

What we call frontier Technologies, Cobras has a set of 17 that we define very broadly in terms of technologies that are changing very rapidly, that are mostly enabling their change and in their capacity to transform economies and society by the because they are digitally driven, they draw a lot of their potency from the fact that computing capacity is accelerating very fast and is much more widely available for for companies and for individuals.

And so we're saying 1 defining characteristic is the rapid expansion.

We estimate that in the 10 years between 2023 and 2033, they would go from 2.5 trillion to over 16 trillion.

That's at 20% growth rate annually.

And this rate of of change and this rate of growth often overcomes the capacity of countries and societies to adapt to, to this change, to understand what is happening and what are the implications for for the future and to react in a way that outcomes are inclusive.

Inclusive, both in terms of the national domestic dimension of change, but also in terms of the international distribution of economic and more generally political power.

That is why the UN is involved in providing policy support, policy analysis to developing countries so that they can adapt their responses, their policies to this very fast changing reality within this set of frontier technologies.

We have been looking at this last year at the one in particular that has been attracting a lot of attention that's artificial intelligence.

From this 17 or I'm on this 17 technologies is the one that is growing the fastest.

We estimate that it will go from representing about 7% of the global market from Fortier Technologies in 2023 to around 30% of this global market for global Frontier Technologies.

This is about a 25 times increase in a period of of 10 years and it would make the market for artificial intelligence about four time four times larger than the market for smartphones.

That gives you an idea of the impressiveness of the market for artificial intelligence.

Now this of course represents huge opportunities for the private sector, but also for the timing of public, public policy objectives.

But at the same time, it involves significant risk of aggravating existing digital and in more general terms development divides.

And that's why we have been looking at the likely trends in artificial intelligence and how national international policy could address them.

One obvious factor that we see and Estorbian already mentioned in his introductory words is that deep digital divides that are immediately translated into divides in artificial intelligence.

I, I will not elaborate too much on, on some of the figures that, that we gave, but just to, to give a few examples, three companies that are very active in the AI field, companies like Apple and BD and Microsoft, each of them have a market capitalisation that it's about the same size as the, as the GDP of the whole of Africa.

1/3 of the world's supercomputers that are critical for the development of artificial intelligence are located in the US Looking more generally at the capital, the the, the funds that are being invested in research and development worldwide, 80% of that is conducted by about two 2500 companies, of which 40% or or half of the funds that are being invested in research and development in, in, in this area is being spent by just 100 companies.

And among those hundred companies, only one is from a developing country.

And that developing country is China.

So yes, a, a final figure, the whole investment between 2022 and 2025 on climate adaptation and climate change was 1/3 of what is being invested in artificial intelligence.

So how can we elaborate the response to to this formidable trend of change in in the world economy and technology and and to the challenges that this sort of presents for developing countries?

One thing is 1 fundamental element is to understand that only a few countries are in the immediate future going to be able to significantly influence the development of artificial intelligence.

But many more countries can influence the way artificial intelligence technologies are deployed in the economy and society.

And different or another element of the framework that can be combined with this distinction between development and and deployment is to understand that there are three fundamental drivers of the growth of artificial intelligence.

One is infrastructure, without which countries cannot deploy artificial intelligence models and those that can participate in the development of the technology cannot develop the the technology without sufficient infrastructure.

And this goes much beyond basic elements like access to electricity, access to water, access to the Internet, but includes other elements, in particular the access to sufficient computing capacity.

The second is that countries need to develop access to good quality, diverse and accessible data without which artificial intelligence models cannot be trained, cannot be tested, cannot be validated.

And the third one is that you need skills, computing skills, but also on very critical data analysis skills so that you can actually know what you're doing with the the data.

And of course, there are very strong synergies between these three drivers of artificial intelligence.

That in particular is the lifeblood of the digital economy and more in particular of the development of artificial intelligence.

Without data that is not biassed, that is complete, that it's accurate, you cannot have a trustworthy inter artificial intelligence and you have a very serious risk of artificial intelligence resulting in deeper divides, in deeper inequalities and entrenching existing inequality.

And that's why I think that improving data governance at the global level is critical.

We need to establish clear rules, responsibilities and oversight mechanisms about how data are collected, how they are used, how they ensure, how they circulate in domestic economies, but also at the international level.

We have currently a fragmentation in the regimes of data governance at the global level and that leads to significant barriers for the participation in the digital economy in general, but more specifically in artificial intelligence and and for the development of inclusive artificial intelligence.

That's why one of the outcomes of the Global Digital Compact that was adopted in September is critically important.

In this sense, it's the mandate to establish a dedicated to working group on principles of data governance at the global level as relevant for development.

This is work that is already started here in Geneva under the auspices of the United Nations Commission on Science and Technology for Development.

That's a body that is serviced by ANTET.

It has met already in the month of May and it will meet again next week.

And they are starting to develop a work programme with the participation of actors from the civil society, from the academia, from the industry, and of course from governments, and not only from developed countries governments.

That's a very important consideration to come up with a set of recommendations for the General Assembly to elaborate more precise regulatory frameworks for data.

Beyond the question of data, of course, there are aspects related to the impact of artificial intelligence on labour.

We have been looking at how labour markets will be affected by the deployment of artificial intelligence and we again see a clear distinction between the consequences for developed and developing countries.

In the case of developing countries, they are less exposed to the impact of artificial intelligence in terms of loss of jobs by automation, but the same time they are also much less exposed to the benefits that would derive from the improvement of labour productivity through the application of artificial intelligence, the capacity of artificial intelligence to enhance the productivity of knowledge workers in particular.

So there is a need for countries, and particularly for developing countries, to adapt their policies, to develop solutions that are adapted to their local infrastructure, to the availability of data that exist in their economies.

And also to come up with models that empower workers in the use of artificial intelligence, that involve workers in the design and the deployment of artificial intelligence so that we keep the worker at the heart of the approaches to implementing artificial intelligence in the workplace.

I will not go into the details of what we recommend countries to adopt us as national policy.

Maybe we can have that at at a second moment in in, in this briefing.

But I would like to say a few words about some recommendations that Antar is making in terms of what can be done.

To strengthen global collaboration around artificial intelligence.

Because there are very few countries that are influencing the development of the technology, but all countries are going to be very, very deeply affected by how this technology moves ahead and how it is implemented.

There are currently about 8 major global efforts in the development of artificial intelligence governance and there is almost 120 countries, 118 I think more precisely, that are not involved in any of these global efforts, while the G7 countries are involved in all of them.

This is an issue that we need to to address seriously because without serious involvement of the developing countries in the development of artificial intelligence framework, we are going to, as I have been saying all along, strengthen existing divides and existing inequality.

There are already some efforts going on on the GDC in in its outcome includes already the beginning of a process for this through the establishment of artificial intelligence governance dialogues and the creation of a scientific panel to advise on AI policy.

But there is a real need to strengthen the participation of developing countries in this process in terms of more practical recommendations, and that is formulating a few of them.

One is the creation of shared infrastructure for the development of artificial intelligence, particularly in terms of computing capacity.

Based on the model of Fern that has been highly successful in providing a common infrastructure for the advancement of research in the field of particle physics.

We think that this is a useful model that can be applied also to artificial intelligence.

We encourage countries to consider new forms of public private partnership for investing in the development of artificial intelligence.

Again, this, like in most areas of the digital economy, this is an area where you cannot do serious progress without the involvement of the private sector and civil society.

And last but not least, to promote the adoption of open data, open science and in more general terms, open innovation for the advancement, advancement of artificial intelligence.

I think I will stop here.

Thank you very much to both.

We are now going to take care of your questions.

We'll start from the room.

[Other language spoken]

You have the floor.

Can you introduce yourself?

[Other language spoken]

This is Mr LAN Tong from China Economic Daily.

It's my great honour to raise this question and thank you all for this area fabulous briefing.

I have two questions if I May 1 is about the upcoming AI for good summit in particular.

The other is about Ankitas portfolio on data governance in general about the AI for good summit.

If I'm right, this year's summit is the first time the multi stakeholder working group established earlier this year and the CSTD is having a meeting on July the 3rd and 4th in conjunction in close proximity with the main conference of the the AI Summit.

So in view of this novelty, what functions or what utilities you are expecting from this debut show of this working group?

And my second question about the Ankita's portfolio of data governance is that as per conventional wisdom, data is has an overarching nature and it has a lot to do with not only trade and development but also other things.

And so yes, I know Ankita has the mandate as per the GA.

But would you please kindly elaborate to us what are the exact portfolio of Ankita when we are looking after from Ankita's perspective on data governance?

Is it particularly with the focus on trade and development or does it have a more overarching horrors and taking into consideration all other aspects of data governance such as security rights, which are discussed under the the whole umbrella?

Thank you very much.

[Other language spoken]

Maybe I will answer the first question on the a meeting of the working group and data and then perhaps Dorian can follow up on the second, second aspect of your question.

The, the fact that the meeting of the working group is happening in close proximity to the A, A for good summit is not exactly an accident, but it's also to some extent and due to to other factors.

Obviously, a lot of the, the majority actually of the members of the working group are going to be present and active in the both in the WSIS high level event and the AI for good summit because they are experts in the field and they are renowned experts in the field.

And that's the reason why they are members.

And logically they these, these there's been a major event in, in the field.

They are logically going to to participate and benefit from exchanges in in the taking place during the summit.

At the same time, these are two very different processes.

The Working Group under the CSTD has a direct mandate from the General Assembly to come up with proposals to advance the dialogue at the political level in the General Assembly on this key question of fundamental principles for data governance.

And and that is a much more structured process.

It's bound by the rules of the UN, even though we try to make it more flexible than a usually it, it's usefully in, in the context of the UN and even though it is not a purely intergovernmental group, as I mentioned, it has a strong participation of non government actors, but it's still, it's part of an policy making process in, in the UN and it's, it will take a considerable time.

We, we are supposed to deliver to the General Assembly these recommendations before the 81st General Assembly, that is in September 2026.

We're probably going to ask for an extension of that.

Because the topics that are to be covered are very, very ambitious and the time available is, is is too short.

But the the discussions are going to be fairly technical and at the same time taking place in a political environment that is particularly complex.

So I think a lot of what you would see in the context of the summit, it's a much more open process would be different from what would be happening in the context of the of the working group.

In any case, we are only beginning what we expect to come as a result of the meeting on 3/4 July.

It's a final agreement on what is going to be the work programme of the group, what format will take the output, the report that the group is going to produce.

And very importantly, it's an agreement on the working methodologies because when you have diplomats in the room, mixing with academics and civil society activist, one important thing is to agree on how they're going to work because they have very, very different approaches to to work.

And it's going to require an effort to adapt the cultures of these groups of people to to be effective working together.

Thank you very much.

Tobian, over to you for the second question.

Yeah, which is a very good question by the way.

Thank you for that.

The question of Anktad's mandate on this compared to the mandate of the CCDS is important to have that distinction because as you know the the Commission on Science Technology for Development is a Commission under Echo Sock in New York, but it is serviced here in Geneva by Anktad.

So that has a broad of course has a broader mandate, mandate than Anktad in its own capacity.

It's true that of course, as you know, Ankit's focus is on trade and development.

But of course from the perspective of sustainable development, which is means that you need to also capture the other aspects of develop and then pure economic development we did in 2019, we we conduct our first report on how to create and capture value in the digital economy.

And in that report we identify that two of the most important driving factors for the evolving digital economy were platformisation and data, which we hadn't really given much attention to in the past.

And it's also an area that has not been given much attention to in the context of the the world stamp and the informational society in the original outcomes.

So this is an area that needed more attention.

We followed up that report in 2021 with a full report devoted only to the question of data flows, cross-border data flows and where we basically map the current situation in the world at the time.

And it hasn't changed that much I must say.

And we, we noted a very fragmented regulatory landscape for data and data flows with the, the very large, the largest parts of the economic economically seen in the world, the US, China and Europe having very different approaches.

And there were differences in how you regulate data and data flows, how you describe data and data flows, very, very large differences in what factors should determine whether data can flow in or out in the countries and so on and so forth.

And we said that in view of the global nature of the Internet, it would be very important for countries to come together and have at least a common understanding of what is important.

So one of the strongest recommendations from our report at the time was exactly to create a new forum for member states to come together and and at least try to come to some common understanding on principles, etcetera.

So we were very pleased to see that Member States in the Global Digital Compact took the decision to establish this new working group.

And it's of course not only Ankh that that's done work here.

You have UNESCO, you have ITU, you have there was a high level body on effective multilateralism that was led by the former Swedish Prime Minister and from the former president of Liberia that also came up with a similar stance, advocating for more global debates on this topic, capturing all the aspects.

We have had this tendency here in Geneva in particular to link this question of data flows to trade.

And because that has been on the agenda in the World Trade Organisation to see how one could regulate it through trade agreements.

But we have also been stressing that it's it goes so far beyond, it goes into human rights, it goes into national security, it goes into many aspects that countries pay a lot of attention to.

So from that perspective, having this dialogue in the Commission on Science Technology for Development is a very appropriate forum because it can breed, go into all these different dimensions.

So we contribute to that dialogue, but the overall work of the CCD goes beyond the pure antimony.

Thank you very much.

I don't see any other question.

Oh, yes, please over to you.

Can you introduce yourself?

[Other language spoken]

Yeah, I, I, I'm, I am soon with warming daily from China.

The first question until the place is a very strong emphasise on economic Policy Research.

I would like to understand what role of the what role the promotion of AI technologies place in global economic development development.

The second question is under tad concerned that artificial intelligence might be a bubble and that it's future development may fall short of expectations, potentially cooling down a key in engine of global economic growth.

And 3rd question, what is what are the mind trends in the future development of artificial intelligence technology where it's focus more on serving enterprises or individuals?

What sectors are more likely to integrate with AI to achieve growth?

And 4th question, do you think that AI currently, currently requires stronger regulations and establish of more rules?

You mentioned issue like increase the waste and electricity consumptions which might be caused by disorderly competition.

However, we all know that excessive regulations my can slow down development.

How do you think the international community should balance developments and regulations?

[Other language spoken]

Who wants to start?

[Other language spoken]

Tobin, that was a very strong set of questions.

We'll do our best to give you our our take on on some or all of them.

You asked about how important it is for us to promote AI policies.

I think as in, in basically in all areas of technological change and that's position is not really to promote the technologies as such.

And AI can be seen as a technology here, but what we want to promote is understanding of potential implications of these technologies.

So we are neutral on whether this is good or bad.

But the question is what kind of policies can turn technologies into something that is in the interest of world development, of economic development, of social development.

And I think the speed of change with regard to artificial intelligence is so high.

And as was mentioned by UNCLE, the the the frontier development is, is located in a very small number of countries in especially the US and China, which also puts pressure on both enterprises of individuals and on governments to understand it in order to know how to regulate it, how to promote it, how to restrict it when it's not in the interest of a country or of the world as a whole.

So what we do promote is more dialogue on this because I think there are very few leaders in the world that fully grasp what is going on in the AI context and also what can be expected to happen in the next 5-10, twenty years.

There are lots of hopes, there are lots of concerns.

And So what we think is important here is also to understand the the underlying drivers of what will determine the outcome of AI.

And again, we come back to the question of data because of control of data will very much influence who will have control of AI.

The the point, the last point you mentioned about regulation of the the environmental implications for instance, is extremely important.

Very few have even given any attention to this until very recently.

It's very important that governments think about how best to regulate the large data centres, for instance, so that the the rapid growth of electricity use and ideally renewable energy does not crowd out other sectors that also crave such renewable energy sources.

So it's a question of putting strict requirements on water use, on electricity use so that it does doesn't have excessive negative effects on the environmental implications.

But since you have just done a very good report related to AI, we'll hand over to the rest to you.

[Other language spoken]

Doug and I, I don't know if I can add much to what you just said, but perhaps to strengthen a couple of points.

We are very, we are very clear that technology is not neutral.

We are neutral about technology, but technology itself is not neutral.

And if governments and societies in general do not have a well thought approach to how they want to handle the development of technology, then the results may be quite negative in in some in some regards, particularly, we have been strengthening in terms of inequality.

We know from history that the outcome of technological development has winners and losers.

And the challenge here is to make sure that we have many winners and not too many losers.

And that's not the way things stand today.

That's not something that we can be sure that will happen.

That's why we call for a number of messes at the national level and the international level.

And that's why we think that it's very important that we support mechanisms for global collaboration around artificial intelligence and data governance is a fundamental one.

But there are others.

And for example, the question of environment is another question that cannot be left exclusively at the domestic domain because it has global repercussions.

Another couple of points on questions that you raised, the question of whether there's a bubble on artificial intelligence, Probably there is an element of unwise investment in artificial intelligence, as has been the case in all major technological changes along history.

It happened with railways.

It happened with electricity, it happened with motor cars in the beginning of 20th century, it happened at the at the end of the 20th century with the Internet, and it's probably happening with artificial intelligence now.

The problem is that while the bubble is happening, it's very difficult to know who is making a bad investment and who is making a very good one.

But what you can do?

Is to try to stay away from hype in the positive or hype in the negative side.

We are not catastrophistic in the analysis of artificial intelligence.

We don't think the concerns about existential threats for humanity are something of an immediate danger it may happen at some point in the future, but it's certainly not the case now.

We do see a lot of opportunities for improvements in many areas of economic and social life, but we also think that unless action is taken, those positive results cannot be taken for granted.

So just try to keep our heads cool.

And the last point on the question of sectors of growth, I think one important consideration about artificial intelligence like most digital technologies is that it's a cross, it's a, it's a multi purpose technology.

It can have bring benefits on in any case changes to virtually all sectors of the economy.

The question is how to magnify the positive impact that it can have, particularly on labour productivity and minimise the risk and the social cost that it can impose.

But it will probably impact all sectors of activity in a very profound way.

[Other language spoken]

Just just to build on what what Anhel said in, in terms of sectors, it's true that it can, it can be applied in different ways across the whole economy, across the whole society.

But in terms of providing the solutions, it's very much concentrated in a relatively small number of very large players.

So that is also why there are concerns that the the market concentration, the market power is tremendously accentuated.

And I think the the the rise of NVIDIA is very illustrative of this.

This single company is the only major player in the global economy that can provide the the solutions that are needed for the chips needed for the most advanced are AI solutions.

So it leads to a very high level of concentration at the global level, which raises also concern about the competition situation and the sharing of the benefits from AI more globally.

Thank you very much.

I see Anyas Pedrejo from AFP online, Can we unmute?

And yes, you have the floor and yes, yes, thank you.

Hi everybody.

Thank you for taking my my question.

It's about something that you say that at the beginning, I think it was her handheld who say that you say that only a few countries are in the immediate future going to be able to sign it frequently influence the development of artificial intelligence.

Could you be more specify how many countries are you talking about and could you name them and explain what do you mean by the fact that there are just a few countries that are able to do that?

Yeah, it's, it's not very clear to me what does it mean exactly.

[Other language spoken]

Perhaps instead of saying it's only two countries, I would say companies from two countries because it's one important change compared to previous moments of deep technological change in history is that the knowledge that is available in the private sector is overwhelmingly bigger than the knowledge that is in the private sector or in the hands of in the public sector.

So yes, it's companies from the China and companies from the US that for the reasons that we have been mentioning before, because of the massive advantage that they enjoy in access to data with which to train and test and validate models.

Because of the concentration in the computing power and the infrastructure that exists in in those economies.

And by the sheer market power that they enjoy.

As a consequence are are driving today change in artificial intelligence and probably for the short term what will happen in the medium to long term, it's of course a different matter, but on current trends we don't see reasons to think that it will change significantly.

Tavian, you want to hide a few words now just.

It is another illustration of this we have done.

We have looked a little bit at where are the most analytical academic papers on AI being produced around the world and in peer reviewed journals.

And interestingly, Google alone produces more such scientific papers than the two leading universities in the US together, which is basically Stanford and MIT.

And if you look at the top 20, most of the institutions that are on that list are either from the United States and China, some from Korea, there are some from Switzerland and some from the UK.

And there are a few, a few other platform companies like like Meta and Microsoft, I believe.

But there is actually among the top 20, there was not one single such institution from the European Union.

So even among the OCD countries, there is a very big difference here between different players when it comes to really making the most frontier analysis and technological development in the area of AI.

To give you a few minutes to think about the next question, if any, I would like to remind you we've issued the World Investment Report 2025 and this year the the, the subject was international investment in digital economy.

So if you want to refer to this, it was very recent.

So we have not invited any members of the of the team of the World Investment team.

It's was just last week, but you can refer to that in the invitation to this press conference.

You had the links to the three reports we were mentioning the technology and innovation report, the Digital Economy report and the World Investment report.

I think it's the end of this press briefing.

Thank you very much for your questions.

It was very interesting and let's see how the AI for good Summit and the with this plus 20 goes and the progress made.

Thank you very much.

[Other language spoken]

UNCTAD Press conference: impact of AI on labor, policies and global equity - 26 June 2025