Session

Summary

This discussion focused on the impact of the first 100 days of Donald Trump’s presidency on technology policy and diplomacy. Experts from various countries debated whether Trump’s actions represent a significant change or continuity in U.S. tech policy. Many participants argued that Trump’s approach marks a disruption, citing increased protectionism, a focus on economic nationalism, and a shift away from multilateralism. The discussion highlighted concerns about the impact on global data flows, AI development, and international trade relationships.

Key points included the potential negative effects on U.S. tech competitiveness due to restrictions on immigration and data sharing. The experts noted a growing tension between the tech industry’s need for global markets and the administration’s protectionist stance. The discussion also touched on the implications for U.S.-China relations and the broader geopolitical landscape.

Participants examined the evolving relationship between tech companies and national governments, with some suggesting a resurgence of state control over the tech sector. The conversation explored the potential consequences for developing countries, particularly in Africa, and the need for regional approaches to technology governance.

The experts debated the future of multilateral cooperation on tech issues, with some expressing concern about the U.S. stepping back from international partnerships. The discussion concluded with reflections on the importance of rebuilding trust in international relations and the potential for education initiatives to bridge divides in the tech sphere.

Keypoints

Major discussion points:

– Whether Trump’s tech policies represent continuity or change from previous administrations

– The impact of Trump’s policies on US-China tech competition and global trade

– Concerns about data governance, AI development, and algorithmic sovereignty

– The role of multilateralism and international cooperation in tech policy

– Regional perspectives on Trump’s tech policies, including views from Europe, Asia, and Africa

The overall purpose of the discussion was to analyze the impact of the first 100 days of Trump’s presidency on technology policy, diplomacy, and global tech dynamics. The participants aimed to provide diverse perspectives from different regions and areas of expertise to assess whether Trump’s approach represents a significant shift or continuity in US tech policy.

The tone of the discussion was largely analytical and academic, with participants offering nuanced views on complex issues. While there were some concerns expressed about the potential negative impacts of Trump’s policies, the overall tone remained measured and objective. Towards the end, there were some more optimistic notes about the potential for education and rebuilding trust to mitigate geopolitical tensions in the tech sphere.

Speakers

Speakers from the provided list:

– Jovan Kurbalija

Role: Director of Diplo

Expertise: Tech diplomacy, digital policy

– Susan Ariel Aaronson

Role: Research professor and director of Digital Trade and Data Governance Hub, Elliott School of International Governance

Expertise: Digital trade, data governance

– Liu Hao

Role: Executive chair of the School of Global Governance at Beijing Institute of Technology

Expertise: Global governance, technology policy

– Katherine Getao

Role: Senior research fellow at Diplo Foundation

Expertise: Cybersecurity, tech diplomacy

– Marilia Maciel

Role: Director of Digital Trade and Economic Security at Diplo Foundation

Expertise: Digital trade, economic security

– Martin Rauchbauer

Role: Co-director and founder of Tech Diplomacy Initiative

Expertise: Tech diplomacy

– Gabriele Mazzini

Role: Research fellow at MIT Media Lab, former architect and lead author of the EU AI Act

Expertise: AI regulation, EU policy

– Michael Kende

Role: Senior advisor for Analysis Mason and chair of the Board of Data Sphere Initiative

Expertise: Data governance, telecommunications

Additional speakers:

– Parminder Yatsingh (mentioned as unable to attend)

Role/Expertise: Not specified

Expanded Summary of Discussion on Trump’s First 100 Days and Tech Policy

This discussion brought together experts from various countries to analyse the impact of Donald Trump’s first 100 days as president on technology policy and diplomacy. The conversation centred on whether Trump’s actions represent a significant change or continuity in U.S. tech policy, with implications for global data flows, AI development, and international trade relationships.

Continuity vs. Change in U.S. Tech Policy

A primary point of contention among the speakers was the extent of change in U.S. tech policy under the Trump administration. Susan Ariel Aaronson argued for significant disruption rather than continuity, highlighting a shift from a trust-based approach to one focused on competition. Martin Rauchbauer supported this view, noting that the changes represent “100 years in 100 days” in terms of the U.S. approach to multilateralism and partnerships. Marilia Maciel concurred, emphasising the erosion of U.S. partnerships and alliances.

However, Liu Hao offered a contrasting perspective, suggesting that while there may be short-term changes, the long-term mission to serve American interests remains unchanged. Michael Kende added that past U.S. government actions had positively impacted internet development, providing context for the current shifts.

Specific Executive Orders and Policies

The discussion highlighted several specific actions taken by the Trump administration, including:

1. Executive orders on H1B visas, affecting tech industry recruitment

2. Changes to data localization policies

3. Shifts in trade policy, moving away from the principle of free flow of data

4. Protectionist measures affecting global tech supply chains

Impact on AI Development and Governance

The speakers expressed significant concerns about AI development and governance under the new administration. Susan Ariel Aaronson pointed out a lack of trust and transparency in U.S. AI development, while Gabriele Mazzini discussed the EU AI Act and its implications for transatlantic relations. Mazzini also mentioned a recent U.S. diplomatic communication to the EU regarding the AI Act implementation, highlighting ongoing tensions.

Liu Hao offered a perspective on China’s approach, focusing on using AI as “intelligent augmentation” to improve efficiency and its application in education. This contrasted with the U.S.-centric discussion and highlighted alternative approaches to AI development.

Katherine Getao brought attention to Africa’s unique challenges, stressing the need for the continent to focus on regional tech policies and data reclamation. This unexpected divergence from the U.S.-China centric discussion highlighted the importance of considering diverse global perspectives in AI governance.

International Trade and Tech Diplomacy

The speakers agreed that Trump’s policies have had a significant impact on international trade and tech diplomacy. Martin Rauchbauer noted that protectionist policies are disrupting global tech supply chains and affecting Silicon Valley. He highlighted the tech industry’s response to these changes, including concerns about restrictions on immigration and data sharing.

Marilia Maciel pointed out a shift away from the principle of free flow of data in U.S. trade policy, while Michael Kende warned of potential retaliation from other countries on data flows. The discussion also touched on concerns about web scraping and data markets, as mentioned by Susan Ariel Aaronson.

Geopolitical Implications

The geopolitical ramifications of Trump’s tech policies were a recurring theme. Susan Ariel Aaronson and Martin Rauchbauer emphasised how U.S.-China tech competition is driving policy changes. Marilia Maciel and Liu Hao both highlighted the erosion of trust in international relations, with Liu Hao emphasising the importance of mutual trust in global tech cooperation.

The discussion also touched on the potential impact on international organizations like the ITU and private sector initiatives such as Elon Musk’s satellite projects.

EU AI Act and Transatlantic Relations

Gabriele Mazzini provided insights into the EU AI Act and its implications for transatlantic relations. He discussed concerns about the implementation of the Act and regulation of generative AI, highlighting the complex dynamics between EU and U.S. approaches to AI governance.

Future Directions and Unresolved Issues

The discussion concluded with reflections on potential ways forward and unresolved issues. Speakers agreed on the importance of education in AI development, with Jovan Kurbalija noting positive developments in U.S. AI education policy.

Several key questions remained unresolved, including how to balance national competitiveness with international cooperation in AI development, the long-term impact of U.S. protectionist policies on global tech innovation, and the future of data governance in light of changing U.S. policies. The discussion also briefly mentioned the CSTD working group on data governance as a potential forum for addressing some of these issues.

In conclusion, the discussion painted a complex picture of the global tech landscape in the wake of Trump’s first 100 days. While there was general agreement on the disruptive nature of recent U.S. policies, speakers offered nuanced perspectives on the implications for AI development, international trade, and geopolitical relations. The conversation highlighted the need for continued dialogue and cooperation to address the challenges and opportunities presented by rapidly evolving technologies in an increasingly interconnected world.

Jovan Kurbalija: Hello, my name is Jovan Kurbalija. I’m director of Diplo, and I would like to welcome you to our session on Trump and tech after 100 days. We are joining, I can see, from the old regions and continents, and we are going to have a really timely discussion exactly on the 100th day of the inauguration of President Trump. And this is obviously a highly important issue, highly controversial, and highly emotional. In order to discuss this issue in as constructive a way, we will follow Diplo’s approach of having informed discussion based on the facts, documents, inclusive discussion, having different perspectives, as you will see from our audience, and impartial, hearing different views on this important topic. In preparation for the session, I wrote a blog post, which already triggers some sort of interesting controversies and debates, basically pointing the framing discussion around the question, is the Trump presidency so far, a change or continuity in overall tech diplomacy of the United States, which dates back to 1875 and the ITU St. Petersburg conference, when basically the US started actively participating in the telecom and tech debates. And this will be, I would say, the underlying question. Do we have a major change, or do we have continuity in the overall US tech, telecommunications, cyber, online, AI? You can use whatever prefix and descriptor as you prefer. Now, we have on the website, and we’ll share the link, we have an interesting AI agent built on the… all executive orders and Trump’s articles written about tech and Trump’s presidency, which you will be able to consult. We’ll share the link in a few minutes. But what matters more than AI agent that you can check anytime is our group of experts, colleagues, panelists who are joining us today. And we have with us Susan Ariel Aaronson, research professor and director of Digital Trade and Data Governance Hub from Elliott School of International Governance, joining us from the center of the debates, physical center from Washington, D.C. Then if we move to the other side of the United States on the Pacific coast, we have Martin Rauchbayer, co-director and founder of Tech Diplomacy Initiative based in San Francisco. And he had to wake up early. It’s now seven o’clock in San Francisco. Thank you, Martin, for joining us. If we move further towards the east or west, we have in Beijing Liu Hao, executive chair of the School of Global Governance at Beijing Institute of Technology. Then in Strasbourg with us is Marilia Maciel, director of Digital Trade and Economic Security at Diplo Foundation and also Brazilian, her origin from Brazil, who will bring us also a bit of the Latin American perspective. Not very far from Marilia is Gabriele Mazzini in Brussels. Gabriele is architect and lead author of the EU AI Act, currently a research fellow at MIT Media Lab. We have not very far from myself here, Michael Candice, senior advisor for Analysis Mason and the chair of the Board of Data Sphere Initiative. And in Nairobi, we have Catherine Getao, senior research fellow at Diplo Foundation based in Nairobi, Kenya, and former Kenyan. tech diplomat, negotiating quite a few cyber security and other agreements. Unfortunately, Parminder Yatsingh cannot join us from Bangalore in India. He had the last minute urgency. Therefore, he sends his best regards. Therefore, as you can see, it’s a very diverse group when it comes to the different perspective, geographical, topical. And we will start our debate with Susan, who is based in Washington, DC, with a few insights. How does it look? Trump and tech after 100 days of the presidency with special focus on AI developments and AI governance. Susan, please go ahead.

Susan Ariel Aaronson: Thank you so much. So I am one of the fortunate few, I’d say, whose grant has not been cut as of yet. But I would say AI-related research is under constant threat in the time of Trump. Moreover, he doesn’t seem to understand the importance of trust and transparency, particularly in the development of data. And I cite as example, when Vice President Vance went to the AI summit, you may all recall what he said, which is basically, the only thing that matters in terms of AI is competitiveness. And unfortunately, I would say that Democrats and Republicans in Washington have a very simplistic vision of what it takes to build competitive AI. And I would say it, like diplomacy, must be built on trust, right, a form of social capital, and it must be built on predictability. Otherwise, people won’t want to use American AI. Moreover, there has been a long history of American companies misusing data. I cite for you as example, Meta slash Facebook’s sharing of data with Cambridge Analytica, which may have played a significant role in the vote for Brexit. So in terms of the United States, unfortunately, just as I would say, chat GPT set up a vision that, you know, AI is going to be incredibly useful, and individuals are going to use AI and it’s going to rock their world, deep seek has led both Democrats and Republicans to have a very paranoid vision that we must be faster, better, etc, than China. And I think that that is unfortunate, because it doesn’t mean better AI. And by better AI, that’s a term, you know, that has no definition. But better AI has to, like any technology, it has to be built on trust. I know I keep saying that five letter word, but I think it’s really important. In terms of priorities, there has been a significant change. Biden’s executive orders set out to make the US government a model of AI governance. And by a model, he meant accountable, open source, with public participation. In my own research, I have found that during the Biden administration, unfortunately, the United States was not, while it was, I think, a model of openness, it was not a model of public participation. The US government, like many governments, and I do comparative work in this regard, solicited public comment. It really did not use public concerns to influence the direction of its AI policies. President Trump has made it clear that he doesn’t care. They don’t care what the public thinks. It’s all about competitiveness. He only listens to the top business leaders. My concerns are that the tariffs are leading to, as obvious, unpredictable business markets, and that’s going to have an impact on competitiveness in AI. Let me give you one example. The economy minister of France recently said that if the United States challenges digital sales taxes as a trade barrier, which it could be, I mean, that’s fair. That’s been long before Trump. But in any case, he would restrict data to U.S. models, which is insane, because if I were France, I would want to have the U.S. have models that include French culture, French science, French language, etc. So that’s a cutting off your head to spite your face. The United States should be working with other countries to develop AI that is trusted, that is trustworthy, that is accountable. And only in that way can we ensure that AI is not misused to surveil people, to undermine their rights, etc. I can wrap it up there.

Jovan Kurbalija: Thank you, Susan, for setting the stage and perspective from your research. And you highlighted quite a few interesting points on the question of competition versus trust, the question of potential blockage on the data as a source or knowledge for development of AI models, which we’ll be hearing later on, how is it seen from other perspectives. But before we move to Europe, first, let’s go to the West Coast. And Martin, how is the atmosphere in the Silicon Valley and the Bay Area? Unmute, just. Please, thank you.

Martin Rauchbauer: Yes, good morning here from San Francisco and Silicon Valley. I think it’s an interesting question to look for the moods in Silicon Valley after 100 years of the Trump presidency. If we take a moment and go back…

Jovan Kurbalija: Martin, Martin, can I just stop? You put the typical flip by saying 100 years, it’s 100 days, because you probably read Marx’s recent book.

Martin Rauchbauer: Yeah, that was a Freudian slip, yeah. Yeah, the first 100 days of the Trump administration is what I meant, and I would like to go back to the beginning of the Trump presidency and maybe look at the kind of expectations that came out of the tech industry or, you know, the tech, I would say, friendly ecosystem here in Silicon Valley. And I would say that at the outset of the Trump presidency, the expectations were largely positive, or at least more positive than during the first Trump presidency. And that had something to do with largely, you know, with the expectation of a kind of deregulatory approach that the US government was going to take. There’s certain ideological affinities of the kind of mindset, although it’s hard to, you know, say that it is unified, but this kind of techno-utilitarian, techno-optimistic atmosphere that is often prevalent in Silicon Valley, of course, has its supporters in the Trump administration. And I think Elon Musk is a very kind of visible symbol for that. So I would say at the outset, by and large, the expectations were positive, very pro-business. the regulatory government that would kind of limit or lift some of the restrictions on technology companies. I think that the atmosphere has changed after 100 days and not years, and that has something to do with what is affecting businesses around the world. I think I think it has been said correctly that businesses thrive not only in an atmosphere of trust, but in an atmosphere of predictability. And I think the disruptions that the tariff discussions and the tariff announcements have caused in terms of business outlooks, in terms of supply chains, if you think of a big behemoth here in Silicon Valley, Apple, that has 90% of its iPhone productions is still in China. What that means for just that one big company, the expectations of large tariff increases, means that there is a disruption, obviously, on the horizon, a potential disruption on the horizon for this particular company. And even if, you know, these disruptions are not going to play out in the most catastrophic ways, I think that there is a general realization among many business leaders that the time has changed in a kind of more systemic way, that tech companies don’t really usually buy into this kind of economic nationalism that some parts of the Trump administration adhere to by the very nature of their business models. They’re transnational, they’re kind of, they expect and hope for global market access. And so I would say after 100 days of the Trump administration, there is a lot of pessimism when it comes to the economic outlook, and I would say some sobering up of some of the positive expectations at the outset. Thank you.

Jovan Kurbalija: Thank you, Martin, for bringing also two aspects of economic interest. One is what you mentioned, development of the hardware or basically industry supply chains, and Apple is a good example, adding to what Susan mentioned to data as basically a key resource for AI development and a threat from France that it may be cut. We now go across the Atlantic Ocean and we go to Brussels to Gabriele, the key architect of the EU-AI Act, and to hear from him, what is the view from Brussels on AI and the first 100 days of Trump’s presidency?

Gabriele Mazzini: Thank you, Jovan, and thank you for having me here with you. Yeah, of course, while I’m in Brussels, of course, I do not speak any longer on behalf of the European Commission, so that, of course, has to be said because I left the Commission last year. So what I’m offering here is, of course, only my personal perspective on that, and let’s say I want to bring back a little bit the experience that I saw firsthand during the actual development of the AI Act, where we could see certainly with the past administration some sort of alignment, because as you may recall, the executive order by the Biden administration, which was referred to by Susan earlier, came out in October 2023, and that was a period when the negotiations of the AI Act were undergoing. And in fact, we could see at the time how, if you want, those two jurisdictions were pretty much aligned, at least in certain respects, in certain sort of principles and ideas. Of course, the executive order was not the AI Act. So it’s a very fundamental difference that an executive order is not a legislative endeavor like it was in the EU. But in both cases, we have seen some sort of alignment in terms of to respond to what was at the time sort of the pressing question about charge CPT and whether this powerful models, the large language models or systems, the frontier models will generate some sort of very risky safety issues that would deserve some attention. And so we could see an alignment, for instance, in shifting in the EU, the regulation from the system level in the AI Act, which was based on the risk-based approach of systems to the actual model. And so in fact, as the audience may already know, the AI Act has introduced a new chapter about general purpose AI models, focusing specifically on the foundation models. And in fact, the Biden administration was including specific provisions on that, on the dual use foundation models. And in fact, there was also an alignment in terms of, let’s say, what elements should be looked at when fixing certain rules. The Biden administration deferred to the flop counts. So the number, the amount of compute used for training the model, that would trigger certain disclosure obligations on the providers of the models, notably disclosing the test reports. And that same approach was imported also in the EU AI Act, because the AI Act regulates foundation models. in two levels, very briefly. One set of rules apply to all foundation models and one set of rules applies to models with systemic risk. And there you can see that one of the criteria that the IARC has introduced to make the regulation switch was, in fact, the flop count. It was a different number, so it was not 10 to 26 as in the Biden administration, but it was 10 to 25. But still, conceptually, that’s what I wanted to underline. There was somewhat an alignment in terms of how to proceed. So, of course, now going fast forward with the Trump administration, that executive order was rescinded. And so in the EU, we still have, of course, a legislative approach and model that is still fundamentally enlarged and inspired by that approach, which now does not exist in the US any longer. And so that’s one aspect that I think it’s useful to consider when thinking about what’s the EU-US relationship is today, looking a bit backwards. But then going forward, just very briefly, not to take too much time in this initial statement, I think, of course, it’s interesting to look at what happened in Paris. So in Paris, first of all, I think that summit was an AI action summit, it was not an AI safety summit. And certainly the kind of tone that I think the Vice President of the United States expressed in his position, his speech, was also in some form, also in different form, reflected from my perspective, also in certain other speeches, like, for instance, that of President von der Leyen. I mean, just looking at it, I was looking at it this morning, you know, the AI Act was mentioned only once. And basically to say, yes, the AI Act has the advantage of creating the common framework in the EU. sort of harmonizing legislation in the EU for one single market, but also she immediately pointed out to the need for simplification, sort of cutting red tape. And so, difficult to speculate how, you know, the change of administration in the US will impact going forward. Actually, very recently, there’s been some reporting, which we can talk about it later, of an actual official demarche that has been sent by the US government to the EU, to actually, among others, request posing implementation of the IAR. So there’s just a very recent reporting yesterday from MLEX. But so, I think there are signs that certain balance is shifting. That’s a bit my initial speculation at this stage. Thank you.

Jovan Kurbalija: Thank you, Gabriele, for this excellent point from Brussels. And you indicated two things. One is basically shift from the difference now across the Atlantic Ocean when it comes to approach. And the second is overall shift, which we noticed in the Paris summit from the spirit of the Bletchley and Seoul previous summits, which was more risk security oriented towards more action. Therefore, it wasn’t just the question of a shift in Washington DC, but I would say general trend that we can discuss later on. Now, so far, we have been mainly with AI. Now, we move a bit towards broader issues. And we have my colleague Marilia Maciel, who’s based in Strasbourg. Marilia, what would be your take on this interplay on the 100 days of Trump’s administration and impact on tech?

Marilia Maciel: Thank you, Jovan. Hello to everyone. I think I’d like to approach this topic more from a geoeconomic perspective. And I think that the first thing to consider is that the digital economy has become simply the economy today. Most economic sectors, they deeply rely on digital technology to function. And I think that this means a few things. The first thing is that traditional economic statecraft, the use of economic means to achieve foreign policy goals is turning into technology statecraft, the use of technology to achieve domestic and foreign policy goals. And over the years, the US has perfected its technology statecraft, mostly to compete with China using instruments like de-risking, export controls, investment screening. And what we see today in this administration is the result of a consistent effort being put in place by different administrations to adopt technology as a weapon of statecraft. And I see a continuity there. We also see a growing interdependence between technology policy and trade policy. So for example, the US president suggested that tariffs on China could be reduced if China approves the sale of TikTok, but China has delayed talks on TikTok because of US tariffs. So we see that there is a need to play a very strong link there. And I think that these changes are also evidenced by a transformation in the US understanding of what economic security is. So if we look at the early days of the internet, when liberalism was the predominant rationale there, economic security meant fostering the global liberal order online. So mostly the role of governments would be to step back and simply pave the way for private actors to expand across borders. Now I believe that we are entering a completely different phase here in which liberalism doesn’t function as an ordering principle of the global economy anymore. And Liberation Day showed that very clearly. So this I think it’s a key discontinuity with previous administrations. Actors in the US and also other actors are seeing digital interdependence as a threat and they are trying to hedge against it. So in this scenario, the goal of economic security has become mainly to strengthen the state, but not with the goal to protect individual rights and strengthen competition, such as was the goal in Europe with the GDPR and the DSM and all the regulatory initiatives. Now the goal is to strengthen the state so it can face these external threats and beat the real or the imaginary adversaries that we are creating today. So this is very clear with the motto that economic security is national security that we have seen in the 2021 US national security strategy. But the problem now is that national security is ultimately about very important threats, which could undermine the existence of the state. So when the national security mindset kicks in, the interest of the market, the duty of accountability between governments and society goes to the back of our thoughts. Markets can still try to push back the state, but they lose leeway and the interests of the state come first. So for this reason, I think that it’s not precise to assume that what we see today in the US is a triumph of the liberal order or alliance between the government and the big tech companies. Some tech companies, they stand to gain when their interests converge with the interests of the state. We have seen recently now a big change in US public procurement rules, which will allow the administration, such as the Pentagon, for instance, to procure ICTs from companies such as OpenAI and Palantir and other big tech companies. But at the same time, the tech companies, they are confronted with a clash between their markets and economic interests and the government’s geopolitical agenda. And we see that in the US. the framework on AI diffusion, which was approved by Biden, but has been upheld by Trump. And this framework will control the way that semiconductors, AI model weights and cloud service for AI will be accessed across the world. That will have a very significant impact in developing countries. Countries such as Brazil, they are sitting down and trying to think how they’re going to have access to AI and to develop their cloud services with these export controls in place. So there is this tension, but I think that overall the tension will be resolved by enforcing the interests of the states. And the last point I would like to make, which makes a link with the content policy, is that I think that because the interests of the states come first, the U.S. has adopted an erratic understanding of freedom of expression lately. Before the Trump administration, there was a discussion on redefining the limits of intermediary liability after Cambridge Analytica, for example, but those that are in favor of more responsibility for platforms, they have lost ground vis-a-vis those that adopt a more absolutist approach to freedom of expression, which weakens the possibility to remove false or misleading content from the internet. But at the same time, we see that some groups such as LGBTQ plus content creators, they say that it has been much more difficult for them to increase visibility online. They receive more flagging. So there is a narrowing down of freedom of expression for some particular groups in the U.S. By and large, I think that it’s a very complicated scenario. But I think that if this neomercantilist view successfully takes root, the danger is that the deeply interdependent digital value chains that we have today will become an obvious choice of weapon to hurt the other side. And that will necessarily bring negative consequences for the digital economy, but also potentially negative consequences for internet openness and for the global technology stack, interdependent stack that we have today.

Jovan Kurbalija: Marilia, and there is now a thread with Susan, Martin, and you, and Gabriele, is a bit of a catch-22 for companies. They rely on global market and the whole success of US tech is basically based on the access to the global market. All the US market is big. And now suddenly they have this, what you call neo-mercantilism, and countries asking, should we rely on them? Or French asking, should we send our data? Or countries asking a reasonable question, if open AI servicing, I don’t know, Pentagon or NSA, should we send our data? There are many issues which are emerging. And I think this catch-22, geoeconomic catch-22, for tech companies is going to be quite, quite big. By the way, you can pose the questions, because we have quite a few speakers. You can pose the questions in the chat. That will be our main interaction. Agreement, disagreements, questions for more information. And welcome to Liu Hao and Catherine Guettao, who were stuck in traffic, but I’m happy that you managed to arrive on time. And we will just move to Geneva, to Michael Kende. And Michael, all of these issues are basically, these big issues, somehow come to Geneva, to WTO, ITU, economic diplomacy, which you are a specialist in. What would be your five minutes take on these reflections from mix of Geneva and the economic aspects of our discussion so far?

Michael Kende: A very interesting question. So, I mean, there’s obviously a lot of… tariffs and export controls. Many of them are likely, first of all, it’s very hard to be precise about what they are. They’ll probably change during this call what exactly the tariffs and export controls are, but it’s likely to not necessarily help the US and backfire to some extent because, you know, if they start putting tariffs on imported chips, that will impact the US companies that import the chips, the companies that use the chips. It’s not clear if the tariffs like on Taiwan have been put in place yet, but it’s unlikely to help US manufacturing in the short run. There was talk of the tariffs put on China would be imposed on smartphones. That’s been removed for now, maybe because people are putting out estimates that if building an iPhone in the US would cost $3,500. And as Martin said, 90% of them now come from China. But I did read that they just imported, Apple just imported 600 tons of iPhones from India to try and get around the tariffs in the short run at least. But there’s also going to be some blowback and some of it will backfire. So the chip exports to China already started to be limited under Biden and now more limits. But this has resulted in Huawei coming up with a more advanced chip for their phones. DeepSeek didn’t have access to the same number of GPUs from NVIDIA that other companies have and apparently was able to train its AI with more. much less, and now they’re working on developing an AI chip that would compete with NVIDIA, so that may backfire. And then one that hasn’t come up much is, apparently, China controls the processing of 92% of rare earth materials that are very important for many high-tech and defense and medical tech. Seven of them so far have been restricted of exports to the U.S., which relies on China for 70% of them, so that’s going to start impacting the manufacturing. So if you take the goals as given to improve U.S. manufacturing and U.S. tech, some of this might backfire. Going to Geneva, obviously the WTO has not really had much impact yet on any of this, that a lot of these trade tariffs are being imposed because of national security exemptions in U.S. trade laws, so they claim that they can impose these either because of the import of fentanyl or other security issues, some of them may be economic, of course. And so far, the WTO, I don’t think, has had much impact. More broadly, the ITU seems not to be impacted yet. I think that the U.S., that people, that countries have to make like four years, not investments, but four-year plans or contributions over four years, so even if they wanted to pull out like with WTO, it would take some time. And there was some discussion that the ITU was exempt because they assigned the spectrum that that Musk’s company needs for the low-Earth orbit satellites. For the rest of international Geneva, I think the impact is going to be more indirect. The huge impact of what has happened with USAID and the other things is more on humanitarian. The, you know, refugees, migration, UNAIDS are all heavily impacted and some of the humanitarian organizations that relied on it. So the Swiss government, the federal and cantonal government is moving their support towards the NGOs that support humanitarian and the organizations that support all of that for good reason. But it means I think there’s going to be less support for NGOs and others that do more tech in Geneva. So I think it’s going to be an indirect impact that there’ll just be less support available as they focus more on the humanitarian for, I think, good reasons. So I’ll pause there. I think it clearly hasn’t played out yet. I think that the USAID one had a big impact on some projects and areas of tech among others. But I think it hasn’t all played out yet. It’s not clear yet if they’re moving down the list and if ITU and others are going to be impacted. But that remains to be seen. So with that, I’ll turn back.

Jovan Kurbalija: Thank you, Michael, for the both economic reflections and the Geneva reflections. Now, Liu Hao, Professor Liu Hao, we are moving to Beijing. China was often mentioned in the reflections and as there is underlying tone on the interplay between Beijing and Washington, D.C. What is your take from the Beijing Institute? Institute of Technology, on the discussion so far and how is this viewed from Beijing. Please go ahead.

Liu Hao: Thank you, Jovan. It’s good. It’s really happy to see now our colleagues online. And well, first, I wanted to share my understanding of the position of President Trump. Some people say he’s unpredictable, but for me, I consider he’s really predictable because he had to make everything clear in his campaign to be the president. And he’s following in the mind of his promise. So he’s now working on what he had mentioned in his campaign. Well, he wanted to be the good president of the United States. Yes, he’s working on that. But whether the result will be really serving for the interests of the United States people, it needed to be evaluated and decided by the American people. So in the short term, or at the very beginning, people may consider, OK, this is really defending or serving for the interests of America. Maybe it is also serving for the interests of American people. But in the long term, we need more time to evaluate the final result. Well, for the reaction to China, we have to say that we are living in an ecosystem. Probably some of our colleagues know that, like playing Tai Chi, when you hate people very heavily, you will also gather the Viking force, which will hit you in another direction. So that’s why recently we see that a lot of actions are taken. taken by the Chinese government and the adaptive changes of the Chinese company. So I see, is that a message to me? Because I didn’t read that message very clearly. You want me to address more on the ITU, right?

Jovan Kurbalija: No, no. This is a question specific on ITU, but maybe you can elaborate a bit more. We will come to this question by Marek. It’s a question for the whole panel, but please go ahead, continue.

Liu Hao: Okay. So this is the Tai Chi theory or the ecosystem theory, which is also quite a part of the Chinese culture. The Chinese people is a group of people that we always wanted to do business. We wanted every people get benefit from the business. So when our counterpart wanted to do a business, we are ready, but if our counterpart wanted to have a war or wanted to fighting with each other, the Chinese people may take the proper action. But the action is not for even to upgrade or to a new level of war or conflict. We are always ready to have a better or have a peace settlement. So that is why in the Chinese character, how to say, the fighting or the weapon is, if you may interpret the Chinese character in the Chinese way, that is a stop fighting. It means the weapon or the fighting. So that is why we believe that such a culture or the philosophy is well implemented, even handling current interaction between the United States and China. China is just a part of this new scenario. So we are not only one of the players. So it is a multi-player game. So the action from the United States, the action from China, the action from Europe, the action from the other part of the world will have interaction. So we wish that China could work together with the other regions of the world so that we may work together to let the world get back to the order that everyone feels comfortable and everyone may get a benefit. Because the current situation is that it seems no one gets a benefit from that. We can see that in the normal people’s lives. We can see that in the performance of the international market. And definitely, we also take the pains at the right time. But even from the first time when President Trump was working as a president, we know that we will not only have peace, we will only not have good business with the United States. So from that time, China is ready for an even worse situation. So the international trade between the U.S. and China has gone down to a scale that also reduces the negative effect to China. So we hope that we can find a solution based on the collective action of the international community so that we can get back to the order that everyone is getting a benefit. But the current situation is not brand new. Because at the very beginning of the world order, every country is serving their own interests. So the globalization is not all the time. Every country serves their own interests. It’s quite natural. But in a global community, we need to know how that works. together how that we make who exists and so that we, everyone, get even better results than your national reserve, than your only serve for your narrow interests. I will stop here, Jovan.

Jovan Kurbalija: Thank you, Hao Liu, for giving the, to paraphrase all the Deng Xiaoping saying when they asked him about the French Revolution. Was it good? And he said that it was too early to pass the comment. Therefore, your answer was in that way. It’s too early to see what will be the impact of Trump’s presidency, not even 100 years, but we need basically maybe 50 or 60 years. Thank you. And we’ll get back to this question by Marek on the ITU, but let’s go to Nairobi. Kat, it’s great to see that you managed to move through the busy streets of Nairobi and you are now by computer. Let us know how does this look from Africa and what is your take on it?

Katherine Getao: Okay. Good day, everyone. I know we’re all in different time zones and thank you for your hard questions, Jovan, which are always very stimulating. Disclaimer, of course, I’m just speaking as an African and Kenyan citizen concerned, so I don’t represent the African Union or any formal body. So, of course, the first 100 days of President Donald Trump have been quite an eye-opener for everyone and a bit of a roller coaster. So I’ll just look at it from a few perspectives. Firstly, his personality. He has not been shy about his contempt for Africa and Africans in general, and this is well recorded. So this has really demonstrated in the In the past, in the last quarter century, our news used to be very much state-controlled or state-mediated, but now Africans are getting their news unfiltered largely from social media. So, they do hear all these statements and they draw their own conclusions. And I see a polarization that comes from reading what Donald Trump says and does, which is reported on social media. Some people approve of him because they see him as a conservative and who follows values which are perhaps closer to the views of many conservative Africans, but others, of course, are appalled by his contentious statements about Africa and Africans. And of course, nobody controls what various people think. So this is also quite an interesting technology-oriented development. The second area is policy. So if I looked at the executive orders, which Jovan, you are so kind to distribute to us, and clearly the words I would use are dominance and protectionism, although I mean it’s good that there’s AI capacity building and other good things there and protection and growth of the infrastructure, which is very good. Now how this will affect Africa? Of course, we’re very, very interested in fintech because that has been a game-changer on our continents and especially in Kenya. So the game is still ongoing as to how this will impact us. Sometimes it impacts us because we sometimes follow the example. But I think it may mean that African innovations may find it much more difficult to globalize because of the protectionist stance. It may also mean that we are unable to access the technology that we need to develop quickly, as I think Marilia pointed out. And I felt that there’s some financial domination. And indeed, I’m hoping to write some papers on this, because this is really very important. The way value and money is being handled around the world and how it’s being regulated has a huge impact on access to goods and services around the world and access to technology. Of course, the way in which Africa, Kenya have been involved in AI has largely been as digital workers. And I think there have been a few scandals there as well. So I’m not sure that that legislation really addresses the people who work in the AI value chain. I don’t think I saw anything there that is really addressing them, unless they start maybe trying to access the IP as they work. It doesn’t really consider that. So I’d say that countries have to really look at this issue. Because the African continent is a very young continent, it’s a tech-savvy continent, and any area of limitation or restriction may impact us, and certainly our governments have to address it, because this is the hope of the future. This is where we hope to make work for our tens of millions of youth around the continent. So we have to think very hard how to respond to this. The third area I’ll mention is precedent, because as various speakers before me have mentioned, some of the actions have been quite authoritarian. Some of them have shown a lack of respect for rules and law, and this is very attractive to autocrats around the world, because America used to be the seat of democracy. So when people see that there can be autocratic action by governments in developed countries, which previously greatly respected principles like free speech and the rule of law, it sort of opens the door for autocrats around the world, and we have a few of them in Africa, to be validated in their actions. This is not very good for their populations. That given, I think the actions of Donald Trump in the last 100 days have been helpful in mind shift around the world. I’m happy that China is looking into now developing high technology, because if we can’t access it from the US, of course we have to access it from somewhere. I hope our governments, the global tech giants are very present in Africa, and I hope that having seen what is happening in the US, that our governments are going to take a more robust view to negotiating with these companies as they profit from our citizens. I think, not that I believe in reciprocating or revenging, but I do believe that people, their eyes should now have been opened as to the risks and exactly what can happen depending on the stance taken by a government that has a lot of power and a lot of technology access, technology that we use. So we’re thinking very hard, the thinkers, how to organize ourselves in this new world. And part of that will be how to reshape the digital future in the light of the new risks that have become very real in this new world. I’ll stop there.

Jovan Kurbalija: Thank you. Thank you, Kate, for this new, interesting element on fintech. I have to admit, I have never thought of this, but obviously, we have M-Pesa and whole cryptocurrency developments in Africa, and we have still to unpack all layers of impact of this decision. Thank you, first of all, for sticking to, let’s say, five-ish minutes. It wasn’t easy, given your wide knowledge on your really top experts in your field. Thank you for that. We’ll go quickly through the questions that we got, and then we’ll go back to Susan to answer these questions or some other questions that we raised in discussion. What is echoing in my so far hearing this are a few aspects. This, again, question, is it continuity of change? Is it a major change or disruption of the US tech policy, which has been going for 150 years, or is it basically a continuity? That’s still a dilemma. I would say that more arguments are for the change, but we’ll be hearing. Marek has always asked a pointed question, and that was the question, Marek, that I asked myself, and we will have the event with people from the ITU. What is the really direct link between US not cutting support for ITU, which is good news, and basically interest of Musk for the low orbit satellites? Maybe Liu Hao could also comment on that in his turn, because he’s also involved in satellites and these issues. Is there a disconnection, or it’s just basically the overall policy? And Michael answered this question, and then Marek, a bit, you may be specific. I didn’t completely decipher this question. Please, more on how improvements of US manufacturing on US oil would trigger even bigger competition. In my view, the tough competition would have been going on anyway. It has many layers, but we will ask our panelists to answer. Then Helen asked a few questions, pointed to Gabriele, about risk and global trade, especially is the UAE going to affect transatlantic economic relations? Maybe Susan and Martin can reflect on that. Also, the question of algorithmic sovereignty, does the Trump administration post-signal shift towards algorithmic sovereignty, rather than interoperability? It’s also an interesting question, also for all panelists, but in particular, Susan, Martin, and Gabriele. AI cryptocurrency, this is an interesting for Martin. Future of AI attaché, Martin leads a tech attaché initiative in Silicon Valley. Do you foresee the role of AI attachés in embassies becoming standard digital diplomacy? So what should be their core skill? Let’s say this is more on tech diplomacy, which will have a chance, but maybe Martin can quickly reflect. Many questions. I hope that LN is not the chatbot AI system generating the questions, as I recently experienced in one of my sessions. Silicon Valley and the state, given your experiences, but questions are good. How do you interpret the evolving alliance between big tech and national governments? That’s basically a lot of questions. Back to Susan, Martin, Gabriele, picking up on a general discussion, but also a few specific questions. Susan, over to you.

Susan Ariel Aaronson: Well, in terms…

Jovan Kurbalija: We can hear you. Go ahead.

Susan Ariel Aaronson: You can? Yes. In terms of trade policy, this is extremely interesting. It’s something I do a lot of research on regarding the free flow data. And you have this weird situation in the United States where, on one hand, everyone was shocked to see all these tech companies allegedly being supportive of the Trump administration, which I do think they’re deeply ambivalent about. My gut is they will use that leverage to press for digital trade agreements that allow for a greater sharing of rules, rather than the patchwork of governance. Marilia and I are probably on the same page of that. But I do think there will be huge differences regarding what is proprietary. There is this weird situation in the United States related to foundation models regarding web scraping, where on one hand, a lot of these companies, and let’s just say Meta and Google as example, and OpenAI, really want to be able to scrape the web and, you know, say this is what we call fair use, which is legitimate for research purposes. But there are a lot of lawsuits saying they have scraped personal information and proprietary information. And to me, that should say to people discontinuity, because there isn’t sufficient governance. You know, yes, GDPR, most companies, most countries have personal data protection and copyright enforcement laws, yet they clearly do not work regarding web scraping. And there is no technical solution, because despite attempts to have robots.txt, really, that is unenforceable. And so you see the situation where firms are, you know, they need more data. What they really need is more accurate, complete and representative data. And there is no country that has policies in place to do that. Another thing I’d like to talk about related to trade agreements relates so not only accurate, complete and representative data sets, for which you do need global trade agreements, but the other thing is, the markets for data are so opaque, it is almost like corruption, we should have counterweights to the collection of data to data broker markets. And only one country and that is China has laws governing data markets and explicit and open data markets in various Chinese cities, but also allows companies to claim data as an asset. There’s been an awful lot of attention to regulating chips, and to regulating compute, much less attention to how do you create an environment to have the data sets. that AI needs or other data-driven sectors. So that’s one thing I’d like to say. In a sense, there is disruption rather than continuity because these rules do not exist. I guess I look at the future and I worry that the United States is a country that’s really lost its confidence. And we have these weird phenomenons. When I was writing my master’s thesis was on competitiveness and it was all about US-China. And I remember going down to Capitol Hill and seeing union workers torching a Toshiba radio and beating up on a Toyota car. This is paranoia. And that’s what we see in the United States right now. The appropriate response is to fund research and to work with our allies. Do I think that will happen? I think there will be a lot of pressure from companies to have more collaboration. What I worry most about, and I’ll shut up after that, is the misuse of AI to survey Americans because of the data collected by Mr. Musk and his pals, but also misuse of AI in defense-related sectors. Thank you.

Jovan Kurbalija: Thank you, Susan. Therefore, change, disruption, not continuity is the underlying message. Few points of interest. Fair use is basically waiting for euthanasia because it’s not fair use if it is used for commercial purposes, not for the research, more or less. And we have a few colleagues from IP community in the room. Maybe they can comment later on on that. It’s an interesting issue. The question of data in China, it’s an interesting development that there is data market in China, probably the only market. And on data, there is one… There will be start of CSTD working group, the title is long, on the data next week in Geneva. We’ll see if they can do something. And for me, one personal observation is that somehow knowledge as a concept more than data was completely diluted and removed from the policy rhetorics after the World Summit on Information Society in Tunis. If you go through the Tunis action plan and commitments and Geneva documents, you had a lot of knowledge, the question of protection of knowledge, and it’s completely removed. It’s only today on data, which is a bit misleading because AI is basically also capturing a lot of knowledge behind the raw data. That’s an interesting issue to be considered by policymakers in the coming years. That’s great. We move then to Martin, few questions to you from Helena on this. One is a question of transatlantic dynamics. You may reflect on the attaches, although we’ll have different sessions on that. But you may just also reflect on that quickly since you lead this process. Over to you.

Martin Rauchbauer: Well, thank you so much. To the overall question of disruption or continuity, I mean, I think there’s certainly, we’ll probably all agree, a certain element of disruption that is characterizing the first 100 days and many of the policies initiatives. The question is, do these disruptions, are they of a systemic nature or are they just an attempt by the US administration to address certain perceived imbalances? And or will these, you know, policies lead to a permanent disruption of the international order? And I found that that really interesting, what Professor Hao Liu said in Beijing, that, you know, we do understand that countries and the US is of course, no exception, act in their own interest. But first of all, is the Trump administration really has it thought these policy measures through? And is it really acting in the US government’s interest? And then, I mean, what kind of system will we end up with? We do need an international system that is based on rules. And that is based on to a certain extent on values as well, otherwise, we all lose. So that’s just a general comment. I just want to point out to a few other factors that of course, have a lot of tech executives and technology scratch their heads. If you think of Silicon Valley, as you know, a region where more than 50% of the workforce is foreign born, we haven’t talked about immigration policy of the US administration, I’ve just recently had a conversation with an immigration lawyer in Palo Alto. If you are, you know, put on more and more restrictions on foreign talent, that’s one of the main, you know, factors of the success story of the US technology sector. And the same goes true if you look at academia, and research and the current struggles of some of the top universities in the country to receive federal funding are for, you know, foundational research that certainly impact. future technological advances. So I think if we look at the kind of international system that is evolving in front of our eyes, I do think that it is obvious that what we call tech diplomacy and that has evolved to a certain extent from the rise of the tech industry in the recent 10 to 15 years, I think that is evident now for many governments that they need to ramp up their diplomatic capacities and introduce, you know, diplomats that are savvy not only in new technologies, but also in the geopolitical implications of these new technologies. I don’t think it’s only a necessity for governments to ramp up those capacities, but also for businesses as well. They need to understand and navigate this new world that is evolving in front of our eyes. And I think to the question of how we’re going to look at this alliance or this relationship between technology companies and governments in general, I think that’s a really fascinating theoretical question, is the re-emergence of the state and taking control of the tech companies in terms of national security and national interest, as Marilia pointed out. The two, I think we see evidence for that as well. So I would say that we are, you know, after these 100 days, looking at a certainly new emerging order that we still need to make. make sense of. I think that Silicon Valley is a very ambivalent place to be with a lot of concerns, particularly for small and medium-sized companies that are really worried, you know, if this traditional success model of innovation leadership is going to hold up. And I think the return of diplomacy, geopolitics to the world of business is very evident when you talk to business leaders here in Silicon Valley.

Jovan Kurbalija: When I was involved in business negotiation, usually diplomats said, let me consult the capital if they need to agree about something. The U.S. delegation was telling that, or U.S. colleague was saying, let me consult the business. And it was an interesting difference. And I don’t know if here we have continuity or disruption when it comes to the constituency, which is going to influence the developments. Now we move first with Gabriele, with this reflection on, let’s say, mainly U.S. transatlantic relations, and then we move closer to Geneva to wrap up with multilateral with Michael, Marilia, Liu Hao, and Kat. That’s Gabriele. You can pick up the question, but there was a question of EU-AI Act and transatlantic relations, which you already hinted upon in your first statement. But please reflect on any issue that you find appropriate.

Gabriele Mazzini: Yeah, I tried to cover as much as I can all those questions. Yeah, first of all, I think, you know, the Act and transatlantic relations… So digital citizenship means a lot of things, for sure. And we can see specifically when it comes to, so the question handed to the recommender system and generative AI. So when it comes to recommender system, I would say maybe upfront, this is not something that actually is in scope of the AI Act in the first place, because it’s more around the Digital Services Act and the Digital Services Policy. And clearly, so it’s still about, you know, the EU having a set of rules that apply to in particular, large American companies. So they’re very large online platforms, they’re very large online search engines. And so that kind of regulation certainly, you know, is enforced, it was enforced even before the AI Act, I think it’s starting to be implemented now. And whether, you know, the Trump administration will react to possible, you know, stronger enforcements of that set of rules on the platforms, you know, the American platforms, it has to be seen. What we have seen in the space of the AI Act and in generative AI, I would say already, and I think it’s what I was mentioning before, that there has been this reporting about the March. So an official diplomatic position sent by the US government to the Commissioner, most likely, you know, other member states, which I’ve not seen, I’ve just seen the reporting on MBLEX. And it’s interesting there that there’s a very, you know, first of all, there’s one part around, you know, posing implementation of the AI Act, which is a fairly, you know, strong demand on the side of the United States. And there is a second set of requests around specifically generative AI. So specifically around this new chapter of general purpose AI or foundation models. And there, according to the reporting, there is a number of specific comments that are put forward, notably, for instance, that the court, some, this court of practice, which are, you know, the primary tool. to ensure compliance with the IAR, they are being worked out by a group of experts under the supervision of the Commission, and some drafts have already been circulated. And, for instance, one point is that some elements of this code of conduct go beyond the legal framework. So, of course, the code is not a law, it’s just a tool to supplement implementation of the law. So, one argument the government seems to make is that, essentially, the code even goes beyond the law, which is, according to their view, a problem. Another point that is being raised is around how will this code apply to downstream providers. As we know, when it comes to generative AI, there is a lot of happening at the level of the model. So, the model can be put on the market by the first provider and then could be fine-tuned later. And some of the risks and benefits of that fine-tuning, of course, cannot fall on the original model provider, which is the focus of the IAR. Other questions relate from the US government around copyright. This was an issue that was discussed. So, whether the approach taken in the IAR around copyright is the right one, considering the complexity of copyright law. I think there is also some interesting comments around transparency, notably how the approach taken in the IAR and in the code of practice will ensure protection of business secrets. Again, this is certainly in line with some of the positions expressed by the US government. And finally, also, there is another set of comments around risk management. So, there is, interestingly, a comment around the fact that some of the risks that the IAR identifies when it comes to the general purpose AI models are somewhat hypothetical and speculative. So, this risk around more long-term, is my as you probably would call it, Jovan, and also around the risk management practices. So how to ensure that those risks can actually be properly managed at a model level as opposed to the application level, notably when you’re dealing with open weights models. And so I think what’s happening there, I think there is quite some indication that this exercise that was started with the IEA Act is now being continued with the Code of Practice, presents some elements of difficulty in actually finding a framework that is easily implementable, provides the needed legal certainty, and ultimately indeed enables that type of innovation to spread. So where are some of those fundamental choices made then, let’s say, make sense. Also when it comes to Gen AI and the question around cryptocurrency, I think it’s really difficult to think that the IEA Act is the solution to all problems of AI. I think this has never been the intention, so it’s not on my end since the beginning. And this should not be what people should think going forward, in the sense that the IEA Act was certainly intended to solve some of the questions around AI development and deployment, but certainly not all questions. And indeed, I think the question, for instance, around cryptocurrency is an example where the deployment of AI in a field like that certainly would probably be better suited to be handled, if at all, if there is any specific need in sectoral legislation. So I think this is also something we need to keep in mind, that the IEA Act, as certainly the first comprehensive attempt to regulate AI, doesn’t mean it will certainly solve its own problems, and it probably shouldn’t, from my point of view.

Jovan Kurbalija: Thank you, Gabriele. for really this detailed analysis of the US-EU dynamics. I’m still, I call it geopolitical schizophrenia, when it was the tech companies were pushing for the EU-AI Act for quite a long time, but then they got cold feet, and now there is cold confusion, which will be interesting for scientists to do research what’s happened, and including interviews with you, since you were there from the very first days in the process. We now move, let’s say, closer to Geneva, we have about 15 minutes, and I suggest that Marilia and Michael, and also Liu Hao and Kat, focus on that multilateral aspect, especially we start with data. Thank you, Peter, and thank you, Marek, for informing us that CSTD already started. It’s an open CSTD working group on data governance. Peter Mayer is a quite active participant, Hungarian official in the CSTD, and now let’s go to Marilia. Marilia, you have many hats, but put your data hat, and we start with that, and then we move to Michael, and we move to Kat, and Liu Hao, and close the geography in this way.

Marilia Maciel: Thank you, Jovan. I’ll do that, but I’ll do that by going back to your question about what predominates, if it’s change or continuity, and I think that for many years, several actors, the BRICS, Europe, they were trying to provoke change, but within the world order that we had, and what we see today is a change of order, and what motivates that is the fact that if we look at different transcripts from different commissions in the US, Congress, for instance, we see that there is a perception that within this order, with these regulations that we have today, the US will not be able to maintain the geopolitical advantage that China is gaining ground, and that the only way for them to escape is to escape the regulations they have created. And that helps us to understand U-turns that are really abrupt, such as the U-turn that they made on the free flow of data, which has been one of the principles of U.S. trade policy. And now they are not supportive of free data flows anymore in terms of making these new agreements with this type of provisions. But I think that if we look, so the ground is shifting below our feet. So I think the net effect will be a change, not continuity, but if we look more granularly, we see that there are some elements of continuity. The idea of economic security is based in three pillars, promote, protect, and partner. Promote is industrial policy. The Trump administration is not giving subsidies to tech industry, but it’s putting in place tariffs. And it sticks to make them repatriate the production of technology back to the U.S. When we think about protection, there’s a lot of the elements that were present in the Biden government that have continued, such as the sanctions, the export control. So there’s a big continuity there guided by this geopolitical competition. I think that the main pillar that we can see change today is the notion of partner that takes us to the multilateralism, the idea that we need to build partnerships to move forward. This has radically changed. And I think that this is the main danger in the policy that is being put in place by Trump, which is this isolationism. The U.S. was using this idea of a world dispute between democracies and autocracies to try to rally the West around the U.S. in order to fight the emergence of autocracies. And this narrative has completely eroded, it’s not there anymore. It has not been substituted by a different narrative. From a political perspective, the U.S. is now alone. And from a practical perspective, it’s not. The alliances that it had are going to be eroded by this lack of trust that has been mentioned by Susan and the alliances that it could make with the developing world and Biden had given signs that he wanted to make more alliances with Latin America, for example, is something that Trump is not really looking at. Two weeks ago, during the meeting of the WSIS Plus 20 review, the US said that they are not going to support the SDGs being part of the WSIS process, which is a major backlash in terms of mainstreaming and development. So this is something that is really changing. And I think that the way that the US approaches multilateralism has radically changed and all the countries are going to reposition themselves in this game. Therefore, I think that’s the main tone will be changed, not continued.

Jovan Kurbalija: Thank you. Thank you, Marilia, for a very interesting aspect, I never thought of it. Within the system, many countries complained for inefficiency, for unfairness, but suddenly when the system is shaking, they say, hey, we complain to the previous system that should have been fixed. Now, we are not particularly happy to go to complete change. It’s an interesting and other research topic to be considered. We have about 10 minutes, we move to Geneva to Michael. And Michael, what would be your take also, you mentioned something on ITU to Marek question, but also on data dynamics, since you’re involved in data, go ahead.

Michael Kende: So just to finish the thought on the ITU, it doesn’t appear that there’s going to be much impact so far. Certainly on the development side, USAID pulled some grants back from the Partner to Connect program at the ITU. But in terms of the radio spectrum, I think from the point of view of Musk and the others, the idea is to get the biggest high possible and then. I don’t see that they would want to pull out of that. I think the pressure is going to be more on the FCC, whose chairman is very Trump supportive, to put it mildly. So I think that’s, that’s going to be the emphasis on the data side. If the focus is on, I mean, AI needs data, if the focus is going to be on competitiveness, I don’t see much change from the US perspective, but other countries like France and the United States are going to have to do a lot more work on that. France may retaliate in ways that may not be helpful to them, as was mentioned, by not allowing data to go through. Other countries already have more protective approaches towards data, either from an economic point of view or a security point of view. So I think those are going to be the challenges for the working group that is starting tomorrow. If I can just take one minute, though, to highlight a positive US government action, which unfortunately took place 30 years ago. The first real internet backbone, internet, was the US National Science Foundation network that connected a lot of supercomputers and academic institutes in the US using the internet protocols starting in the 80s. It had a restricted commercial traffic, which was what made it difficult for the small ISPs, and 30 years ago today, decommissioned that backbone and allowed full commercial use, full commercial access to the internet, use of the internet, leading to all of the, a lot of the changes we see today in terms of internet exchange points, investments and other things. So I think that’s something positive that unfortunately 30 years ago, but it was a positive development of the US government that I think it’s always good to have something positive.

Jovan Kurbalija: Thank you. Michael, let me be positive on the same side, but it’s seven days ago, I think. There is one executive order on AI education and apprenticeship, which was, I think there are very, very good ideas. And what was also positive development, China also started their education policy. And maybe through education, we can avoid this geopolitical competition, because education is always, always good. Not always, but in many cases good, it increases awareness. And I suggest that you consult the executive order from the 25th of April, if I’m correct, which is a very substantively solid of strengthening capacity of schools, AI apprenticeship and other things. Thank you, Michael. We move now to Kate. Kate, we have about six minutes and we will need to hear from you and to Liu Hao. What would be your wrap up? So many threads we have here and you laid out many issues. Give us your reflective thought on continuity and change or anything else.

Katherine Getao: Okay, thank you very much. I think there’s not much time, so I’ll just say two things. Well, when I see the dialogue this year, 2025, it looks as if Africa is not really on the rest of the world’s agenda or radar. There’s very little discussion about Africa. So I think if I was to talk about multilateralism, I would just focus on regionalism. Africa needs to now implement its free trade area. We need to create data policies that are coherent with our situation. And that is we focus on capture, openness and use, because that’s the problem we have. Maybe, you know, we’ve tended to focus on privacy because we’re copying others, but the problem is very different. The other word is I would… that is reclaiming, because a lot of African data lies outside Africa. So on the issue of infrastructure, of course, that is very, very important for us. And Elon Musk had some success, especially in Kenya, with his star network, because people saw it as an independent, less government controlled form of infrastructure. So I don’t know if his company becomes less successful, and he’s unable to maintain what he has already put in place. Of course, that will affect Africa, because the investment was there. Individuals had to pay around between 300 and 500 US dollars, which is a big chunk of their income to set up his infrastructure. So review, regionalize, and have the right attitude towards data. Thank you.

Jovan Kurbalija: Great review, regionalize, and right attitude to data. And we have now Liu Hao with the concluding reflections from Beijing, and on many, many threads you can, I’m sure, reflect. But there are a few things, maybe something quickly on ITU, since you’re involved in satellite issues as well. And this question of motivation of Mr. Musk, and the question of education, maybe that could be an interesting aspect, or anything else, although time is not on our side. Liu Hao.

Liu Hao: Okay, for the education, we are using AI for education, but we are interpreting AI in a different way. We don’t believe that there are artificial intelligence. We consider that we are not having AI. We have IA, which is Intelligent Augmentation, or Intelligent Agents. So we are using that technology to help our professors. to learn more quickly, more efficiently. We are serving them to let them have a better life. So this is an augmentation of these technologies. We consider that we may establish a mutual trust and let the world to be better. Well, regarding roles of the data or roles of technology regulation, I consider that it’s the same. So now we are giving too much emphasis on the competition, on the security, on the sovereignty. All of that is because we are lacking of the mutual trust. It’s the same as the world is now fragmented. We don’t trust each other. And without such a trust, we cannot work together. And the regulation in different countries or the region is like a countermeasure for the mistrust you have encountered. So I hope that in the future we may give more emphasis and spend more efforts to let the mutual trust to be re-established so that we can deregulate, let the data, let the other technology to have free movement so that we may continue to work together and serve for the science, technology and application. So at that time, we may run faster, let the science, technology serve the people and let it serve more people. I’m still positive. Well, regarding the change or the unchanged of the policy of different administration, it depends on whether you are looking in the short term or the long term. For the long term, no change. For the short term, you may have a slightly change. Well, it also depends on you’re looking at the form or the mission. The mission is still the same. The mission is to serve the American people, serving the people who elected you. But for the formation or for the models of the regulation, especially for the executive order, it definitely needs to change to let people consider that this is a new government, this is a new administration. but in the long term for the mission that serves all the people who elect you is still the same. Thank you. I think I didn’t take too much time.

Jovan Kurbalija: Fantastic. How Liu, you were so efficient, extremely efficient. Thank you for observing. We have one minute more. Time to thank you all, to thank Susan, Martin, Gabriele, Marilia, Michel, Hao, and Michael and Kate for great inputs, great comments. And I would conclude with a few dates. Michael mentioned the great date of the internet, which started, expansion started 30 years ago. We are discussing 100 days of President Trump and Martin made the Freudian slip of 100 years. And let’s see what will happen in the 100 years. We won’t be around, but maybe somebody will listen to this recording and say how right or how wrong we were on this discussion on continuity and change. Once more, thank you for our panelists and thank you for all who followed us today. All the best.

Agreement points

Disruption in US tech policy

Speakers

– Susan Ariel Aaronson
– Martin Rauchbauer
– Marilia Maciel

Arguments

Disruption rather than continuity in US tech policy

Shift from trust-based to competition-focused approach

Change in US approach to multilateralism and partnerships

Summary

These speakers agree that there is a significant disruption in US tech policy, moving away from previous approaches and towards a more competitive and isolationist stance.

Impact on global data flows

Speakers

– Susan Ariel Aaronson
– Marilia Maciel
– Michael Kende

Arguments

Lack of trust and transparency in US AI development

Shift away from free flow of data principle in US trade policy

Potential retaliation from other countries on data flows

Summary

These speakers highlight concerns about changes in data flow policies and their potential impact on global AI development and trade relations.

Similar viewpoints

Both speakers emphasize how the US-China tech competition is leading to protectionist policies that are disrupting global tech supply chains and business models.

Speakers

– Susan Ariel Aaronson
– Martin Rauchbauer

Arguments

US-China tech competition driving policy changes

Protectionist policies disrupting global tech supply chains

Both speakers highlight the erosion of trust and partnerships in global tech cooperation, with Marilia focusing on US alliances and Liu Hao emphasizing the broader lack of mutual trust.

Speakers

– Marilia Maciel
– Liu Hao

Arguments

Erosion of US partnerships and alliances

Lack of mutual trust hampering global tech cooperation

Unexpected consensus

Importance of education in AI development

Speakers

– Jovan Kurbalija
– Liu Hao

Arguments

Positive developments in AI education policy

Focus on using AI as “intelligent augmentation” in China

Explanation

Despite representing different perspectives, both speakers emphasize the importance of education in AI development, with Jovan highlighting recent US policy and Liu Hao discussing China’s approach to AI in education.

Overall assessment

Summary

The main areas of agreement include the disruption in US tech policy, concerns about global data flows, and the impact of US-China competition on tech supply chains and diplomacy.

Consensus level

There is a moderate level of consensus among speakers on the disruptive nature of recent US tech policies and their global implications. This consensus suggests a shared concern about the future of international tech cooperation and trade relations, which could influence future policy discussions and diplomatic efforts in the tech sector.

Different viewpoints

Continuity vs change in US tech policy

Speakers

– Susan Ariel Aaronson
– Liu Hao

Arguments

Disruption rather than continuity in US tech policy

Mission to serve American interests remains unchanged long-term

Summary

Susan argues for significant disruption in US tech policy, while Liu Hao suggests that the long-term mission remains unchanged despite short-term changes.

Approach to AI development

Speakers

– Susan Ariel Aaronson
– Liu Hao

Arguments

Lack of trust and transparency in US AI development

Focus on using AI as “intelligent augmentation” in China

Summary

Susan criticizes the lack of trust and transparency in US AI development, while Liu Hao presents China’s approach of using AI as ‘intelligent augmentation’ to improve efficiency.

Unexpected differences

Role of data in AI development

Speakers

– Susan Ariel Aaronson
– Katherine Getao

Arguments

Lack of trust and transparency in US AI development

Need for Africa to focus on regional tech policies and data reclamation

Explanation

While most speakers focused on US-China dynamics, Katherine Getao unexpectedly highlighted Africa’s need for data reclamation and regional policies, presenting a different perspective on data governance in AI development.

Overall assessment

Summary

The main areas of disagreement revolve around the extent of change in US tech policy, approaches to AI development, and the role of different regions in shaping global tech governance.

Disagreement level

The level of disagreement is moderate, with speakers generally acknowledging changes but differing in their interpretations of the extent and implications of these changes. This suggests a complex and nuanced understanding of the current state of global tech policy, which may lead to diverse approaches in addressing challenges in AI development and international tech diplomacy.

Partial agreements

Similar viewpoints

Both speakers emphasize how the US-China tech competition is leading to protectionist policies that are disrupting global tech supply chains and business models.

Speakers

– Susan Ariel Aaronson
– Martin Rauchbauer

Arguments

US-China tech competition driving policy changes

Protectionist policies disrupting global tech supply chains

Both speakers highlight the erosion of trust and partnerships in global tech cooperation, with Marilia focusing on US alliances and Liu Hao emphasizing the broader lack of mutual trust.

Speakers

– Marilia Maciel
– Liu Hao

Arguments

Erosion of US partnerships and alliances

Lack of mutual trust hampering global tech cooperation

Key takeaways

There is more disruption than continuity in US tech policy under Trump’s first 100 days

The US approach has shifted from trust-based to competition-focused, especially regarding AI

There is a significant change in the US approach to multilateralism and international partnerships

Protectionist policies are disrupting global tech supply chains and data flows

There is growing concern over AI governance, particularly regarding generative AI and foundation models

The lack of mutual trust is hampering global tech cooperation and data sharing

Africa is largely absent from global tech policy discussions and needs to focus on regional policies

Resolutions and action items

None identified

Unresolved issues

How to balance national competitiveness with international cooperation in AI development

The impact of US protectionist policies on global tech innovation and supply chains

The future of data governance and cross-border data flows in light of changing US policies

How to address the lack of trust in international tech cooperation

The role of Africa in global tech policy discussions and development

Suggested compromises

Focus on education and capacity building in AI as a potential area of international cooperation

Develop regional tech policies and data governance frameworks, particularly in Africa

Reframe AI as ‘Intelligent Augmentation’ to focus on human-centric development

The appropriate response is to fund research and to work with our allies. Do I think that will happen? I think there will be a lot of pressure from companies to have more collaboration. What I worry most about, and I’ll shut up after that, is the misuse of AI to survey Americans because of the data collected by Mr. Musk and his pals, but also misuse of AI in defense-related sectors.

Speaker

Susan Ariel Aaronson

Reason

This comment highlights the tension between the need for international collaboration in AI research and the potential misuse of AI for surveillance and military purposes. It raises important questions about balancing innovation with ethical concerns.

Impact

This comment shifted the discussion towards considering the broader implications of AI development beyond just economic competitiveness, introducing ethical and security dimensions.

China is just a part of this new scenario. So we are not only one of the players. So it is a multi-player game. So the action from the United States, the action from China, the action from Europe, the action from the other part of the world will have interaction.

Speaker

Liu Hao

Reason

This comment provides a more nuanced view of the global AI landscape, moving beyond the US-China binary often discussed. It emphasizes the interconnected nature of global tech development.

Impact

This perspective broadened the conversation to consider the roles and impacts of other global players in AI development, not just the US and China.

The net effect will be a change, not continuity, but if we look more granularly, we see that there are some elements of continuity. The idea of economic security is based in three pillars, promote, protect, and partner.

Speaker

Marilia Maciel

Reason

This comment offers a balanced view of the changes in US tech policy, acknowledging both continuity and change. It provides a framework for understanding the shifts in policy.

Impact

This comment helped to synthesize various perspectives shared earlier in the discussion, providing a more structured way to think about the changes in US tech policy.

Africa needs to now implement its free trade area. We need to create data policies that are coherent with our situation. And that is we focus on capture, openness and use, because that’s the problem we have.

Speaker

Katherine Getao

Reason

This comment brings attention to the unique challenges and priorities of Africa in the global AI and data landscape, highlighting the need for context-specific approaches.

Impact

This comment broadened the geographical scope of the discussion, bringing in perspectives from regions that had been less represented in the conversation up to that point.

Overall assessment

These key comments shaped the discussion by broadening its scope beyond just US-China relations and economic competitiveness. They introduced ethical considerations, highlighted the interconnected nature of global AI development, provided frameworks for understanding policy shifts, and brought attention to perspectives from less-discussed regions like Africa. This resulted in a more nuanced, multi-faceted discussion that considered the complex global implications of AI development and tech policies.

How will the potential blockage of data as a source or knowledge for AI model development impact different regions?

Speaker

Jovan Kurbalija

Explanation

This is important to understand the global implications of data restrictions on AI development

What is the direct link between US not cutting support for ITU and Musk’s interest in low orbit satellites?

Speaker

Marek

Explanation

This explores the potential connection between US policy decisions and private sector interests in space technology

How will improvements of US manufacturing and US oil trigger even bigger competition?

Speaker

Marek

Explanation

This examines the potential economic and competitive impacts of US domestic policy changes

Is the EU AI Act going to affect transatlantic economic relations?

Speaker

Helen

Explanation

This explores the potential impact of EU regulations on international trade and cooperation

Does the Trump administration post-signal shift towards algorithmic sovereignty, rather than interoperability?

Speaker

Helen

Explanation

This examines potential changes in US policy approach to AI governance and international cooperation

Do you foresee the role of AI attachés in embassies becoming standard digital diplomacy? What should be their core skills?

Speaker

Helen

Explanation

This explores the future of diplomatic roles in the context of AI and technology

How do you interpret the evolving alliance between big tech and national governments?

Speaker

Helen

Explanation

This examines the changing relationship between private sector technology companies and government entities

How can we create an environment to have the data sets that AI needs or other data-driven sectors?

Speaker

Susan Ariel Aaronson

Explanation

This explores policy and regulatory approaches to support AI development while addressing data governance concerns