Éthique, technologies numériques et IA : perspectives de l'Asie du Sud-Est, avec Elina Noor

ANJA KASPERSEN: In some of your writing you state that when looking at the Southeast Asia context there are two sobering realities to reckon with—ideological agnosticism and a resistance to picking sides geopolitically. You opine that there needs to be a greater appreciation of the Southeast Asian context and perspectives. I would like to ask: In your view what are the uniquely Southeast Asian perspectives and experiences that ought to be understood in international discussions in technological standards and rule setting?

ELINA NOOR: I think the overarching theme that you hear in refrains from political leaders in Southeast Asia is not wanting to be caught in the middle of great powers, whether it is in the political or technological realms. So this idea that countries need to choose between one power or another, between vendors from one country or another, doesn't sit very well in a region as diverse as Southeast Asia.

Secondly, the priority that I mentioned earlier in Southeast Asia is to leverage on everything digital and everything technological for the economy of Southeast Asia, so a lot of government efforts and industry efforts are geared toward that. Again, there is not this focus on competing technologically with other countries or even other powers. Therefore the ambition and motivations that drive a lot of technological innovation and a lot of technological and political agendas are very, very different from what you see in countries like the United States or even China, for that matter.

These are some of the prevailing themes that pervade a lot of the tech thinking in Southeast Asia, but I also believe there are missing perspectives in Southeast Asia that I try to draw in some of my work. One of those missing pieces is the fact that in a region as diverse as Southeast Asia we are not thinking about, for example, the indigenous communities and how they view tech, how they view data, and whether they are included or not included in data sets that go toward training algorithms. We don't often think about minority communities and their languages, whether they are represented or not in the race for all things tech in Southeast Asia. While I think the focus to grow the economy in Southeast Asia through tech is a noble one, there are key elements that are missing from the consideration of policymakers in the region.

ANJA KASPERSEN: You recently launched a report called "Raising Standards: Data and Artificial Intelligence in Southeast Asia." Can you share with our listeners the main findings of this report?

ELINA NOOR: One of the main findings of the report is that there is this huge exuberance, hunger, and excitement for everything tech. The acronyms abound in Southeast Asia—IoT (Internet of Things), AI, 4II, you name it, and we have it in the region. But a lot of that focus is very narrowly driven toward the economy, so a lot of the policies, a lot of the strategies, and the framing in general is geared toward commercialization of technology in order to economically benefit the population of Southeast Asia.

There is not an accompanying focus on what for lack of a better term I call the "social justice" aspect of technology. I alluded to some of these—how we account for the linguistic, ethnic, and cultural diversity of Southeast Asian populations, for example, and how we address prevailing inequities in the economy through tech. There are some superficial references to this in the policies of the different governments in Southeast Asia, but there isn't a deep reflection on how to first of all address these imbalances in each of these societies of Southeast Asian countries but also how we project some of our own principles, how we view things, and our perspectives on tech coming organically from our historical experiences in the region as young post-colonial nations, and how we present that to the world at international forums like the United Nations, which you are so familiar with, Anja. The title of the report, "Raising Standards," is about raising standards internally within each of these Southeast Asian countries but also raising Southeast Asian standards to the world at some of these international forums on technological standards, norms, and rules.

ANJA KASPERSEN: For those of our listeners, Elina, who may not be familiar with the Southeast Asian landscape, I think particularly the role of the Association of Southeast Asian Nations (ASEAN), which is now a fairly mature organization that has been pioneering a variety of efforts to bring together common policies and issues. But of course as you said it is a very diverse region.

ELINA NOOR: ASEAN comprises ten Southeast Asian nations, all of which are politically diverse, so there are different forms of government. You have democracies, you have a monarchy, and you have communist governments in power. We are ideologically "agnostic" as I call it. There isn't so much a focus on the form of governments in these Southeast Asian countries.

Each of these Southeast Asian countries is also very, very diverse linguistically. For example, one of the 13 states in my home country of Malaysia has about 30 different languages or sublanguages, and that is only one state in one country. If you go to a country like Myanmar, I think there are about 135 officially recognized ethnicities. You go to a country like Indonesia with tens of thousands of islands, and you will see how very, very distinct and diverse each of these countries is. If we all get our economies in shape, all ten of us would constitute I think the fourth largest economy in the world.

On the Internet e-commerce grew at an exponential rate even during the pandemic, perhaps unsurprisingly because everyone was in lockdown and so there were a lot more new users and there was a lot more shopping going on, a lot more reliance on the super-apps in Southeast Asia that deliver food, bring transportation to so many people, and create so many jobs. I think there were some 30 to 40 million people who joined the Internet for the first time ever during the lockdown. There is just so much bullishness about e-commerce in Southeast Asia from governments, from industry, and even from consumers themselves. So you have this perfect trifecta of hunger for all things tech and e-commerce and shopping online, fintech, you name it. Agriculture? There is a tech suffix to it. Education has a tech suffix to it. There is just a lot of hope that is anchored on technology in order to drive this economic force for the region.

It is already a very robust region economically. I think average gross domestic product rates are projected to be between 5 to 6 percent even post or in the middle of this continuing pandemic. So the outlook is very, very positive, whether it is in physical economic terms or in virtual economic terms.

ANJA KASPERSEN: And the policy maturity that is so necessary to complement that kind of appetite from the users?

ELINA NOOR: That is a good point. Policy aspirations are great. Policy maturity is still quite checkered in the region. You have economies, for example, Malaysia and Indonesia to a certain extent, which are trying to get their policies on the digital economy off the ground. Many of these economies also have complementary AI strategies and policies, even if details are missing from many of these blueprints.

You also have a country like Vietnam, which is often viewed as a very centralized economy, but it is very plugged into the international system with obligations, for example, the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), which obliges participating countries to liberalize their data-driven economies. Vietnam, Indonesia, and Southeast Asia are often viewed as these economies that are more inclined toward data localization, but if you look at the reality of what is going on, the governments in these countries are trying to open up their economies in order for them to be a little more "data liberal," if you will, to attract more investment and economic opportunities in that digital space.

Also, you have so many different religions in each of these Southeast Asian countries. We talk a lot about this diversity in the ASEAN region, but we don't reflect on what it all means in digital space and how we represent that digitally for our own societies but also at international forums like the United Nations.

ANJA KASPERSEN: To your point of the immense diversity, to think that we can develop systems and language models that will cater in favor of preserving that diversity and recognizing the immense challenges of creating inclusive AI models, what are your thoughts on this and what does that look like when you come to highly diverse countries like Southeast Asia represents?

ELINA NOOR: I am so glad you brought that up. I am a huge fan of your work, Anja, and that of Carnegie Council, and I know you have done incredible work on the ethics of some of these issues that you bring up.

In Southeast Asia one of the things that we often overlook, even though we tout it, is this diversity that you talked about. I think what is very reflective of some of the on-the-ground challenges that the ASEAN countries face themselves. For the young post-colonial countries that are only about 60 or 70 years old, a lot of the focus of government is in nation-building. There isn't this mirror reflection of how to continue that nation-building process in the technological space, reflecting a nation as it were of many, many different communities and societies as one in the digital space. So while there is this race for all things economic, there isn't this parallel race for how to build a nation, whether it's Indonesia, Myanmar, Malaysia, or Singapore, and how to reflect that online and to the international community in the tech space as well.

You brought up the whole effort at being inclusive with natural language processing (NLP), for example. But I think what we should also be aware of, particularly in Southeast Asia, is that there might also be an inclination to not be included in some of these data sets, whether it is NLP or others. I say this because I have been looking at—and I am sure you are familiar with this, Anja—some of the images communities elsewhere in Oceania, the Pacific, and in North America, where they have been approached to have their languages included in data sets to be more reflective of an inclusive and diverse world, but some indigenous communities have basically said: "Thanks but no thanks. We will keep the language to ourselves."

That is something sacred to them that needs to be taken into account by the different governments, and because of the natural historical tensions between indigenous communities and governments in many, many countries, even in Southeast Asia, I think there has to be a deeper consideration of whether or not to include some of these languages or identity markers of particular communities if they don't wish to be included. That has to also be respected or considered at the very least, and that is something that we haven't in Southeast Asia begun to even scratch the surface of.

What I understand is that because of some of these structural inequalities there is a mistrust between indigenous communities in particular and the governments of the countries that they now live in about the handling of their data and everything that represents them. So there is this idea of continued oppression, of imperialism, and of colonialism that prevails in many of these efforts to include these communities. Those efforts may be well-intentioned, but because of those historical inequalities that linger those intentions are not translated accordingly.

I think there has to be a reckoning of how to handle some of these challenges. There is an indigenous data sovereignty movement that you might be aware of and that is worth exploring in different contexts. It may not work in Southeast Asia. It may not work in other indigenous communities as well as it does, for example, in North America, but I think it is important to at least try to connect some of these parallel or similar challenges that marginalized communities, whether indigenous or not, face in order for us to arrive at a satisfactory response or satisfactory responses as the case may be, depending on the country and society.

What I have heard from industry is a recasting of this idea of data sovereignty. Obviously there are different motivations from among different stakeholders, but industry's perspective is that, "Well, data sovereignty shouldn't just be tied to the idea of geography and territoriality," so it is possible to maintain a sense of sovereignty even if that data is stored somewhere else. Obviously again, different perspectives from different stakeholders. I think it is worth having more of a conversation about this, particularly in diverse regions like Southeast Asia.

ANJA KASPERSEN: If you allow me, I would like to talk a little bit about you, being at the frontlines of some of the most defining issues of our time, not least defining issues also for the region that you have grown up in. You said initially that you were from Malaysia. Looking at your incredible professional journey and your accolades so far, I was curious as to what exactly inspired your specific research interests?

ELINA NOOR: It is kind of a twisted journey. My background is in international law, so I guess I have always been interested in these ideals of justice, social justice, and international justice, but I think early on in my career I found myself going the "conventional route," if you will, going into think tanks, working in the policy space, and being more concerned about the geopolitics of the world. I think as I have become older I have also become a little more skeptical and a little more cynical about the machinations of the world and great-power politics.

I found myself being a little more reflective over the past couple of years, particularly over COVID-19, when we were all locked down and a lot of time to reflect on many things. I found myself reflecting on power imbalances in the United States, especially when the George Floyd murder happened—that obviously coincided with the pandemic—but also power imbalances in the world, which I guess is very obvious to many people, but I think I started thinking a little more about it and how that spills over into the technological domain as well.

I thought a lot about the decolonization of tech and whether it is at all possible to address some of these power imbalances, rethinking what it means to be from "the Global South," and being really a part of the "Majority World," as some scholars have begun to call it, but not being represented proportionately at international discussions and organizations in the same way.

I thought a lot about what it meant for us to be a market and a marketplace, for us to be consumers and users of technology rather than producers and innovators at scale in the same way as some of the major powers are, and what that means for our own digital future and the creation of our identity in that digital space over the next 20 to 50 years. I guess I have become a little more philosophical and a little more reflective about tech and the politics of tech.

ANJA KASPERSEN: You said something really interesting, Elina: "Is it possible to decolonize tech?" There is an assumption in there that tech is an extension of colonial thinking. Can you elaborate on that?

ELINA NOOR: That's a great catch, Anja. That was maybe a Freudian slip on my part. Thank you for calling me out on that.

Yes, I guess there is a colonial underpinning of tech because if you look at the power balance or imbalance a lot of that power is situated in Silicon Valley, increasingly in Shenzhen, and in the capitals of major powers, but a lot of that power is drawn from the collection of data from elsewhere and the extension of commercialization and the markets in countries that provide that data, and a lot of those countries are situated in Asia, Africa, and Latin America. So there is a structural continuity of some of this oppression.

Not to exaggerate, but it is colonialism in a sense because if you look at census and data collection in the colonial period that all went to the concretization of power of many of these colonial countries in the past, and that continues today, just in a very different form. I have spoken to historian friends who say that what we are seeing is nothing new. It is just a different way of doing things that were done in the past. That may mean that we need to try to be a little more mindful about how we view tech and how we cast and frame technology in what we assume to be a changing world but may not at all be a changing world.

ANJA KASPERSEN: I think with that you said that the pandemic has given us all time to reflect, and maybe with age we do become a little more—I wouldn't call it "cynical"—realistic perhaps about what is going on, which brings me to something I have seen you talk and write about, which is how current approaches actually risk fragmenting and polarizing the current technology and AI landscape rather than fostering models of governance and collaborative opportunities.

ELINA NOOR: Yes. I think we have seen that most recently over the last few years where I think in Southeast Asia in particular because there is so much focus on the promise of tech there were questions about what competing initiatives like the Trump administration's Clean Network initiative, for example, versus China's model of national governance of tech that was counter-proposed after the Clean Network initiative was launched.

I think a lot of us in Southeast Asia were wondering what that all meant for interoperability of networks of vendors and of suppliers in the region particularly because in Southeast Asia government and industry users are so familiar with a mishmash of all sorts of different systems, networks, and vendors. So this idea that there could be a "splinternet," a fragment of how the Internet works, of what systems may or may not be interoperable with each other gave pause to many policymakers because we in Southeast Asia tend to want to be friends with everyone. We try to live this ideal of unicorns and fairies in international relations, and that extends to the digital domain as well.

We are still struggling in the region with what it all means for us, particularly since a lot of our connectivity projects, including digital connectivity projects, span different countries and go across borders. I don't think we have the answer just yet. Even 5G decisions tend to be recast in political terms when reported in the media when actually there are more technical reasons that underpin a lot of these decisions. So we don't have the answers just yet, but at least I think our policymakers are beginning to be a little more aware of some of these political pulls and pushes in the technological space.

ANJA KASPERSEN: We mentioned the digital space. Some of your work over several years now has been on finding ways to strengthen the stability of cyberspace, and you have served on various boards and commissions and written about this in depth. Can you speak a bit more about the field in general but also some of your work and some key insights you have drawn, both on what you have seen going on but also where you see us heading in this domain, again maybe also bringing those very unique Southeast Asian perspectives, mindful of course that you don't speak on behalf of the region and that the region is very diverse?

ELINA NOOR: Thank you for that caveat. Yes, I don't speak for all these [laughter].

One of the things that has been heartening to me just observing developments over the past decade and discussions on cyberspace, for example, is the fact that ASEAN, which is typically criticized as moving at glacial speeds—I think even glaciers move faster than ASEAN sometimes. You may know this yourself, Anja.

ANJA KASPERSEN: Mostly because of global warming . . .

ELINA NOOR: That's a good point.

ASEAN has actually moved pretty rapidly relatively speaking on issues like norms of responsible state behavior in cyberspace over the past few years. From 2018 onward, propelled by the leadership of Singapore, we have seen a lot of capacity-building and training programs that have taken place in the ASEAN region to try to get government officials up to speed on what is expected of them in order for these countries to be considered "responsible" actors in cyberspace.

On the other hand, I think there is also a fair bit of awareness of what this all means—norms, rules, and principles inside this space—in Southeast Asia, particularly for countries that want to be a part of the cyber ecosystem but do not quite yet have the capability and capacity to be active participants in cyberspace. These countries are still grappling with [whether] to classify particular industries as critical infrastructure or not.

At the United Nations Group of Governmental Experts on Information Security (UNGGE), within that process, you only have Indonesia, Malaysia, and Singapore as members of the UNGGE, so that is only three countries out of the ten ASEAN countries that have had exposure discussing and negotiating some of the texts surrounding norms of responsible state behavior inside this space. Three out of ten is a start. It is not a great start. I think there needs to be a fair bit of exchange, information sharing, and knowledge sharing.

Also, a considerable amount of organic growth that needs to take place within the ASEAN region about what it means to represent ASEAN principles and perspectives at some of these discussions, whether it is at the UNGGE or the UN Open-Ended Working Group on Developments in the Field of Information and Telecommunications in the Context of International Security on some of these issues related to the governance framework of cyberspace.

ANJA KASPERSEN: Do you feel that cyberthreats are well understood in the Southeast Asian landscape?

ELINA NOOR: I think it differs from country to country. There are countries that view cyberthreats very narrowly, and so they are concerned about things like phishing and cybercrime, and then there are other countries that are slightly more technologically mature that are more aware of and have been exposed to cyberthreats that involve geopolitical actors, so state-sponsored advanced persistent threats, for example. These threats are not shared equally. The threat landscape is not seen the same way across the board amongst all ten ASEAN countries, and that is only natural given the differing levels of capacity and capability, but increasingly there is some effort among ASEAN Member States to share these experiences in order for all countries to have some sort of baseline from which to work from.

For example, a lot of Southeast Asian and other countries in the Majority World contribute a fair amount of peacekeeping forces, and there are implications for digital threats in the peace and security sense because when you have peacekeepers in foreign lands there is also an obligation to protect the personal information of these peacekeepers or the data that you are handling in conflict zones, for example. I have not seen any discussions publicly or not in my region at least, in Southeast Asia, about what this all means for peace and security, for peacekeeping forces, and contributing forces from amongst Southeast Asian countries. That is one angle that I think is often overlooked.

The other angle of course is that, as you mentioned, digital threats are not confined just to the digital space. I know you said not to confine it to Southeast Asia, but I have seen this in Southeast Asia, particularly with the South China Sea conflict. There is often an assumption that kinetic conflicts are kinetic, and digital conflicts are digital, but as you pointed out there is crossover or spillover between the two increasingly. With regard to the South China Sea we have seen this going back to about 2009 when the threat landscape evolved from just information collection and surveillance to something a little more, which is operations below the threshold of armed conflict.

There are a lot of these developments that are taking place below the surface that I don't think are being adequately considered, and maybe that is deliberate because countries are a little sensitive about talking about some of these things, but there is space and there are opportunities for these discussions to take place, whether it is at the [military level] or even at the political level, behind closed doors. I don't know that for many countries in Southeast Asia, because of their focus on economy and the digital space, that there is an inclination to talk about peace and security issues, digital threats, or to have these conversations at some of the ministerial meetings that go on in the region, but I think there definitely should be more of a public discussion on this.

ANJA KASPERSEN: As someone who has been in that field for some time as well I am often pondering when I get this question if we are really talking about digitalization of conflicts or are we talking about securitization of digital goods. I am sure you have been thinking about this as well, and I would be very curious to hear your views because it fundamentally changes how we approach it.

Going back to your very thoughtful approaches of challenging how we frame issues, how do we stay honest and humble around this new set of vectors that we need to consider when we talk about international relations and international security? Very often we retroactively apply concepts that we know to something new, but we could also do ourselves a disfavor in that approach, and when we speak about the digitalization of threats, the digitalization of conflicts, I sometimes feel that is what we are doing. Instead maybe it would behoove us to talk about the over-securitization of what is happening in the digital space.

But maybe I am wrong and maybe you disagree with me. I would be very curious to hear your views.

ELINA NOOR: No, not at all. I completely agree. I have found myself caught in that binary, and I wonder if there are other approaches that we are missing out on in trying to frame some of these issues. Again with the region I am most familiar with, Southeast Asia, there is a comfort in staying away from the securitization of the digital space because Southeast Asian governments just don't like to delve into issues of peace and security even when the governments see that they have to.

ANJA KASPERSEN: It is a core pillar for ASEAN.

ELINA NOOR: It is, but as you know and as the whole world has criticized ASEAN for this, that core pillar is often sidestepped because it is so difficult. That is why there is such a huge focus on the digital economy instead. It is less controversial in a way.

But I wonder if there is something in that sidestepping that might be worth picking apart a little more because as you mentioned we have one or two choices now. I wonder if we are doing ourselves a disfavor as you said by not considering other approaches to framing some of these challenges that we find ourselves in. Maybe there is a middle ground or middle grounds. I don't know the answer to that just yet. I am also pondering it myself.

Anja, please, if you arrive at an answer, please just ping me and I would be very, very happy to hear about it.

ANJA KASPERSEN: I think we are all going to be pondering that for some time.

ELINA NOOR: I think the one main question I have is—and you alluded to this much earlier, Anja: Why are we optimizing? Why are we digitalizing? I think this is a core question that we have not reflected on and pondered on in a thoughtful manner because we are in this race to optimize and digitalize everything, but why are we doing it? Are we doing it just for the economy? It is fine if that is the case, but I think there should be something more.

I tried to bring this out in the report that you mentioned that I just released. The goal of optimization or digitalization, or optimization through digitalization should be to empower and advance the human potential, and it should be to address social inequities, to minimize them in the future. By focusing on the economy or the commercial aspects of it I think we are only just getting to an interim stage of where we want to be. It might not even be an interim stage.

I think that the goal of optimization through digitalization should be to empower the human being and human potential, and this is core to the issue of ethics in AI and ethics in tech, and if we are not answering that question satisfactorily, then we are not going to arrive at the "right" solutions in any meaningful way.

ANJA KASPERSEN: Elina, you alluded earlier to the issue of ethics, which is of course a very important one, and also the fact that ethics is often seen differently or can be seen differently depending on the context you are looking at, the region. I am curious, based on your work what are you seeing from your side? Are there unique ethical perspectives from Southeast Asia?

ELINA NOOR: I think it is very convenient now to have a set of ready principles of ethics. A lot of countries in Southeast Asia are borrowing, adopting, and adapting. You have the usual buzzwords—transparency and accountability—but how you translate those linguistically into local languages is one thing. How you interpret some of these concepts is another thing. So there is a translation issue on many different levels.

I have seen this in the course of researching my report, where concepts like "privacy" translate very, very differently in different communities that have more of a communal sense of belonging. The idea of individual privacy does not translate linguistically, conceptually, or culturally in communities that basically take it upon themselves to know what their neighbors are doing because there is a sense of care and of responsibility for the well-being of those neighbors. Those of us who are more attuned to the concept of individual privacy might find that a little intrusive, but there is a different perspective of privacy that is shared that should be taken into account by some of these other people and communities.

Likewise, how do you translate "accountability" and "transparency" into a local dialect, for example, in Indonesia if there isn't that sense of what it means whether linguistically or conceptually in the way that we understand it at international levels?

On the one hand, I think there are seven baselines of ethics that we share—the idea that you should do no harm through technology, that there should be an agency that is represented in the use of and design of technology—but on the other hand there are these different levels of what these ethical principles mean at the ground level that need to be unpacked a little more.

ANJA KASPERSEN: Thank you so much, Elina, for sharing your time and impressive regional and topical expertise with us. This has been a very insightful and engaging conversation.

Thank you to our listeners for tuning in, and a special thanks to the team at the Carnegie Council for hosting and producing this podcast. For the latest content on ethics in international affairs follow us on social media at @carnegiecouncil. My name is Anja Kaspersen. I hope we earned the privilege of your time.