IGF 2023 - Day 3 - WS #405 Internet Fragmentation: Perspectives & Collaboration - RAW

>> AVRI DORIA: Welcome. We seem to have passed the bottom of the hour. I guess we should start. Welcome to this session on, what is the exact title? "Internet Fragmentation: Perspectives & Collaboration".

So that's a good clue as to where we are heading. It's been an interesting topic to watch being talked about this week. Lots of opinions. Lots of definitions. The beginnings of a new framework for how to understand it and discuss it, which is still growing and still being thought of.

So there's really a lot of really interesting people, knowledgeable people on this round-table and we really want to get a discussion going of all the people around the table.

And also, as we move on, all of you that are sitting back here. So anybody that is going to want to talk is going to have to come up. There are only a few microphones so you will have to come up and get a microphone when you want to talk. But anyhow.

I want to welcome you all. I really want to get started. As opposed to me saying a lot more because you all have a lot more to say. So, Elena Plexida from ICANN, would you like to start us off with a view?

>> ELENA PLEXIDA: I'm on. Hello, everyone.

Thank you, Avri. Yes, I can kick off with some more opinions and some more definitions, which I'm sure you will be very happy to hear.

So as you know I work for ICANN, representing a technical organisation. And therefore, if you will, what will come here with does come from the technical world.

I think that to talk about internet fragmentation, once we start by defining, by singling out what it is that makes the internet single. What it is that makes it one. Because as we all know, when I was having this discussion with Avri at breakfast, the internet anyway, is not one. Maybe it's of different networks, all kinds of different communication infrastructure. What is it that binds it together to what we call today the global internet? This is none other than the unique identifiers, domain names and the name space, the IP addresses and internet protocols alongside.

Okay, I'm not a technical person, so I think of it as some sort of common technical languages that all devices speak. They can find each other on the network. And the emphasis being on the word unique. It's this uniqueness that gives us the global internet. As long as different networks and devices connected use the same unique set of identifiers we have one internet.

And that's of course ICANN's mission. One internet. To ensure a stable and secure operation of the internet. Unique identifiers, we will do that with the sibling organisations, RIR's. That's it, a fragmented internet would go against everything ICANN stands for. Everything ICANN was created to do.

But what is, what would be internet fragmentation? At the content level there are already limitationsment content is not available to everyone, everywhere. That's been happening for years and even desirable in some cases, think of parental controls. Of course not desirable in other cases. That's not fragmentation, that's limitation to content a user can access. User experience. Fragmentation, if you will. But it's not fragmentation, it's actually confusing because it was told to be a little inflammatory, it is to my mind dangerous to keep referring to content level limitations as internet fragmentation. Because people leave the discussion with the impression the internet is already fragmented. That can become if you will a self-fulfilling prophecy. Talking my own experience, speaking with parliamentarians about legislation that would create this or that. When we go to that point they were saying it's already fragmented, so, why would we, what is there. That's why I say it can become a self-fulfilling prophecy. The internet is still there. It's not broken. Fragmentation would be if we take example of the postal service, would be if the postal service stops being there. If I tell my post man that I don't want to receive letters from Avri, that's not internet fragmentation, it's just that I don't want a specific part of the content.

Fragmentation is when you don't have interoperability. Is the internet fragmented today? No. Absolutely not. Can it be fragmented? Yes, I think it might. It might.

Alternative name spaces. If we have that, the uniqueness is gone. A second root of the internet, the uniqueness is gone. I will not go into the technical side of it. First of all, I'm not technical. And second and most importantly, because I think that although fragmenting the internet is a technical issue, it will not come if it comes from the technical world, it will come from the political world.

By accident, the latter the accident being what concerns me the most.

The million dollar question, if you will, is will the global internet survive a fragmented world? You know, we live in a world that is not the same as it used to be before. There is a lot of po litization around a number of issuesment we start to see this politicalization around the internet universe as well. That can be dangerous for the very global nature of the internet. I will stop here.

>> AVRI DORIA: Thank you, Elena. The next person I have is Jennifer Chung from dot Asia to give her definitions, et cetera.

Thank you.

>> JENNIFER CHUNG: Thank you, Avri. My name is Jennifer Chung, I work for dot Asia, a registry operator of dot Asia domain.

I guess from my point of view, whether or not I can add to the definition or add to the controversy of this discussion that hopefully we will have here is, dot Asia is a registry. We of course sit on the application layer of the technical part of the internet. And I think perhaps we are quite clear on the fact that what technical fragmentation might be. If we start with the baseline, if we are looking at where are we starting this assessment from? If the assumption the primary benefits of the core features is to be able to have universal connectivity and to have the interoperability, I'm really bad with this word, between these consenting devices then I think there's that very baseline that we can agree upon. I don't think a lot of people are really confused about it, or would argue against this part.

I think where we are coming from now, I think many different definitions try to bucket fragmentation into different categories. I see a lot of papers and research and also opinions saying that first there's a fragmentation of the technical layer which hopefully is not controversial. Secondly there's fragmentation of the user experience, or more on the end user or how we experience how we navigate. And thirdly, fragmentation, mainly on the policy level. Which is more governed by, you know, places where decision making is made there. Or in governments where there's decision making on policy, regulations, legislation that could aim to destabilize or could fragment the internet as we see it.

I think what is really important that we should also remember is what isn't fragmentation. I think the word fragmentation is now used, it is very important to use this word, but if we use this word to describe every single thing that is different, I think it behooves us to actually pull back and realise, no, this is actually something that is good for the development of the internet.

One example I would like to bring out from the dot Asia point of view is, a lot of people see internationalized domain names as, hey what's going on here? Could there be a threat of fragmentation? Is this already a fragmentation. I would like to posit to say internationalized domain names which means domain names that can be seen in scripts such as erdo or the Hans script, Chinese, Korean or Japanese, use the Hans scripts. These scripts allow you to see the domain name in the native script. The threat here really if this is not implemented well, then we have the possibility, or the danger of having a fragmented internet. Not the fact that we are implementing this becomes a fragmentation of the internet. So that's the one thing I would really like to bring up first. The second thing is when we are looking at a different part of fragmentation, when we are looking at, now I'm talking more about the policy level. When we are looking at where we are sitting right now at the internet --governance forum, we are talking about these things but when we look at bodies that decide regulations, upcoming legislation, what we really have to remember is that, when these actions and legislations aim at this content at user layer and this causes internet fragmentation it also threatens the technical layer, because then the implementation comes down,  internet shut downs come from policy level. Or when people ask certain bodies to shut down portions of the internet. Those are the geopolitical concerns and pressures we have to resist when we talk about fragmentation as well. I think I want to end a little bit more at least for my first intervention to mitigate these risks really requires first of all a lot of conversation and coordination. But also not duplicating all this conversation into different silos where nobody is talking to each other and not quite getting the part where we need to coordinate well. So I will stop right here.

>> AVRI DORIA: Thank you. It was interesting to me to hear IDN included in the list of possible fragmentation. So thank you for bringing that up. I will be interested to hear more about the whole notion of implementation being something that could cause that.

The next person on the list is Timea Suto, ICC basis.

>> TIMEA SUTO: Yes, thank you, international Chamber of Commerce. For those who don't know us, he  we are representatives of global industry, we have over 45 million members across 70 countries across all sizes and industries.

What does fragmentation mean to me from this perspective and to us? I have to react to what the others have said before, because I think that's the whole point of this conversation.

On that baseline layer that Elena you were talking about, is the internet fragmented? No, it's not. It's working.

But attempts have been made and were sort of successful to disconnect and prove that it can. It really can. And I think I need to agree with Jennifer, your last point there, that certain pressures that come, not at the technical layer, but at the top of all of that, at the content layer, at the data layer, at the policy layer, governance layer have very real impacts on that technical layer, the internet way, this network of networks.

And I think disregarding that and saying the internet works it's not fragmented, is putting our heads in the sand. Because there's real dangers of the internet fragmenting if we buy into the fact that we can fragment the top of it, because it's really for it to go down. That is my, maybe a bit controversial view, but I think we cannot disregard this. Especially when we are at forums like this and others that don't have the technical expertise, maybe this forum has because it's multistakeholder, but other forums that make decisions at the political layers, the content layers and the policy governance layers might not have all that really technical background. So it's easy first of all for them to confuse things. And secondly, thinking if it can be done at the top, why not do it elsewhere. Those are dangerous questions for us to ask. To bring it back to my official talking points, for us what matters here is the digital economy that was built on top of the internet and digital technologies and everything the internet enables. I will talk about digital economy, it's not just about GDP or the bottom lines of business.

But the society, the development goals, the growth what personally, for communities and for economies that was fueled by the internet. And that really depends for us on the ability to move data across borders. To make sure that data supports global trade, information exchange, commerce, healthcare, medicine, research, everything that it is built on the top of data being able to cross borders.

There are barriers to that, data flows are real examples of internet fragmentation. Maybe I don't have a better word to call it. We can put that charge to the audience if you have better ways to call it. But if barriers to data flows coming from various concerns, concerns mostly about trust on the internet, whether it is, I don't trust my data to go outside my region because privacy protections are not the same. Or the IP protections are not the same. Or the consumer protections are not the same. Or just because I think I can create more value by keeping it here and not letting others share it, access it, process it, I think these are very dangerous thoughts and thoughts to data localisation and fragmentation in this layer. I think, first of all, hamper all the benefits of the internet, even if it works technically, the benefits don't come.

And it's not a user choice, right? It's not Avri saying I don't want to receive letters from you. I cannot receive letters from you because others have made that choice for me. And also another question you might want to delve into later. Thank you.

>> AVRI DORIA: You are all starting to answer the question and we are having the discussion. All these people talking about not sending me letters is going to get sad. Anyhow, next I would like to go to Javier Pallero, consultant digital rights tech and culture. Javier is remote?

>> Javier Pallero: Thank you for the opportunity for being there. I am connecting from Argentina, hi to everyone there.

So let me talk with my attempt at responding to this very difficult and specific question.

What I would say is that, I have to agree with all that was said before, I think this is a very complex issue that starts with a very specific definition that is technical, right? And listening to Jen, I have to rewrite thinking about what makes the internet one. Which is unique identifier protocols and common language that is spoken.

But one of the aspects I would like to bring up from civil society perspective is not only what makes the internet unique, or one, but also why that happens? What is the reason to be that the internet has, at least for most of us as users. That's actually the ability for us to be able to communicate and to connect with everyone, right?

So that's, I think what is at the core of this confusion. Or this idea, right? That for example, politicians would say the internet is already fragmented right, there's this perception the reason the internet has to be, has been changing fast. It has become more closed, more seemingly or perceived, in perception it has become more disconnected. More unable to provide that sensation of connection and you know, the ability to express yourself without borders and to access information and so on. I would dare to go a bit further and say that it is actually not a confusion, this idea, you know, that intertwines the political application, technical protocol levels of the discussion. It's not a confusion. It's something that happens because the thing that goes across all of these dimensions is the reason of the internet to be, right? And the reason to be of the internet is for connection and for various technology that enables the enjoyment of rights and so on.

That apparent confusion is actually a part of the problem. I know also asked the last speaker before me Timea said many of these situations, these decisions in the policy level or the application level as well, when a private company becomes a dominant actor in one area of internet services for example. All of that affecting somehow technical decisions, right? So for example, politics can mandate shut downs or data retention, or national gateways, right?

But also certain companies, for example, can insert more influence into certain protocols for example. A key example that comes to mind, the DRM protocols that have been added to the W3C discussions. Many of that comes from private parties, not necessarily governments, right?

Also another example when it comes to government that goes beyond the extreme example of shut downs could be the censorship attempts that are done through mandating changes to the DN S.cerevisiae vers or putting pressure on the DNS servers.

So all of that is a way of saying this dimension, even if it's not part of the technical specific concrete definition of fragmentation which I share is more of a technical specific discussion that maybe can be benefited by being correctly framed and limited. But all of these aspects I've been mentioning are also important. They may not be part of the definition, but they are part of the problem. And part of the perception that has to go with the idea we have about the internet and how we think we want to use it.

I think when it comes to working on this, we will have to make a big effort to make a distinction about the different dimensions. Maybe focus on some of them. Like the protocols one. Because the other ones tend to have their own areas of discussion, right? The ones about censorship of applications or bad policies, right? All of those are properly covered, let's say by some other actors, activity, discussion, regulation, civil society actors who are active in that but this area there's not much engagement. And maybe that's where a more narrow definition of the issue could be of service, just to inspire more attention on the underrepresented dimension, if you may. But the fact everything is important and should be considered. So I would stop there for the initial intervention and thank you again for the invitation and opportunity to be there virtually. Thank you.

>> AVRI DORIA: Thank you, Javier. And thank you for saying it was virtual when I said remote. That was an old-fashioned word we aren't supposed to use any more. So the virtual or online. So I really appreciate the correction.

The next person I would like to go to in this initial set of discussions is Nishigata-san. Thank you. Now that you have a microphone, please.

>> Good afternoon, everybody. My name is, thank you for the kind introduction, my name is Nobu Nish, work ministry of internal affairs. Thank you for coming.

And we do appreciate everybody's participation and contribution which made this event good. Very good. Thank you very much, again.

So getting back to the point, internet fragmentation. Since I'm the governmental  --  I can write legislation though, not the code. But you know, it's confusing matter to me, right?

Just some people, it's many, many different definitions. We need a definition before starting talk, right? But I tried my best today.

And then maybe, you know, this is part of my job following what is happening in the internet every day. I do recognise some, or maybe most of the recent issues, just following up on the category of the internet fragmentation.

And then, you know, refrain from put names of countries but I recognise there's frustration in the tech community, different kinds of fragmentation. From the government perspective, I understand the frustration against particularly against government intervention and it's forceful type of fragmentation, for example, like internet shut down during the election period. And then, however, though, like you know, from the government perspective, this is not the job of our ministry but some other part of the government, we have to do some jobs particularly for our public safety perspective.

Particularly within the border, the government has, internet is global, this is great. But on the other hand the border matters to the government, you know. Not only public safety but the government may make some actions that frustrates you guys in the tech community. And for the sake of the other high-level policy agenda for economic development or national safety situations.

Maybe this is a discussion point what the government can do. Understand some communities hate even single government intervention into the government, we understand though, we have to be accountable for these actions and in Japan's case, fortunately though, we don't see these severe cases yet. And of course government of Japan respect open and free internet. You can see  --  or like Japan chairs the G7 meetings this year. And agree to support the DFI our Chair's leadership. Japan government get together and finish our Day 0 session (?)

So I understand the government action in general may frustrate the internet people. But on the other hand it isn't only the internet people that get frustrated. The government also sometimes gets frustrated. I would say at least not satisfied. There are some issues. For example, though, there are issues, I would say like regarding the fragmentation maybe like user interface type of fragmentation.

--

He just committed to our effort to maintain open, free internet and the reason is we shouldn't leave anyone behind.

From the benefit of the internet from the 34 years. Thank you.

>> AVRI DORIA: Thank you, it's been an interesting set of comments it started very low with a very precise. We added nuances as we sort of moved up the scale. And then it's almost flowered to the point of anything that interferes with an open and free internet can perhaps be seen as a fragmentation.

So that is a very good representation of sort of the blossoming of this conversation. The blossoming of the differences that many of us have. I would like to call on some other folks. We have really an amazing number of folks around this table that have probably good things to say. And dig a little deeper. Perhaps there will be other nuances and other extensions that will get added. But also dig a little deeper into some of what's been said. Next I would like to go to Ala G Gembo, a member of parliament.

>> Thank you for the introduction. The Vice Chair of African parliamentary network on internet governance. I think we are discussing a very important topic. That is confusing some people. Because of definition. But I quite agree that anything that actually interferes with the free flow of the internet, you can call it internet fragmentation. I'm a law maker but I came from the tech community.

As we try to have a stable and integrated internet, this fragmentations whether at technical or government level or business level, because these are the three areas you can see the fragmentation may happen. We may have an issue at the level of legislation. Legislators don't want to legislate anything that is ambiguous. We want to be very clear what we are trying to legislate. The internet, you don't want to open any legislation that would actually hamper or stifle innovation.

Again, when you have fragments of the internet like this, little islands, that actually not talking (?) regularly or you know, optimally, we may have some issue. The causes could be political by governments but the side of legislation I think there could be an issue here. We are trying to have a free for all and trying to stimulate legislation across the world. That's why you see we have African parliamentarians we have some from parliament and we have been talking what can we do together to make sure we have a safe, secure and integrated internet.

So bringing these splinter groups causes more problems. Because we already have issues in terms of legislation. Now bringing actually more divisions in this area will cause us more problems. I think we really need to discuss to see what we can do together to answer that we leave it the way it is and make sure we promote a secure connectivity. We can work together as legislators or policy makers to ensure we can streamline what we do.

Now you know, you just mentioned sometimes about internet shut down. But personally I would like to call it internet disruption. Because what is happening right now is, they are trying to, you know, stop particular application from running the entire internet. That is disruption. You aren't shutting down the internet completely but stopping particular application from running. I think all of those things something we really need to look at to see that disruption stop. This would propel more than. It's better we leave legislation the way it is and support more that we are free flow and we also have integrated and we also have it more secure for everybody to live. Thank you.

>> AVRI DORIA: Thank you for basically taking it a little further into differentiating the different behaviours and the different problems. And that perhaps is helpful. Next I would like to go to Tomawaki global communications in Tokyo. Thank you, everybody for helping play the game with pass the microphones. Please.

>> Thank you, my name is Tomawaki, I'm an academic based in Tokyo.

The way I thought about this issue is when is splinternet bad, or bad enough? I kind of agree, or I kind of resonate with the idea that democratic or politically motivated splinternet is the one we should get most concerned about.

But then maybe we should be aware of the fact that democracy, even in some of the most democratic countries is, these days, a challenge to an extent or another.

I think things like war against terrorism, or majors against rioting or civil unrest. Those things are not that foreign to some of the most democratic countries. And I'm sure that some level of internet regulation is desired by governments of those countries. So I don't  --  and also let me add one more thing. That having a free and open internet, I am, in principle, I tend to think that's a good thing that's a condition for better society. But also I think these days a lot of questions are asked how good it is or is this enough to bring about good changes? Or sometimes as many people have already mentioned, it causes really serious adverse effects. In light of those things I think we have to think carefully how to proceed in a way. Because I think it's not really so simple as to say oh, only certain countries are programmatic and these countries are more pro-freedom, pro-unified internet. Because I think upon closer inspection, even in those countries which are pro-democracy, pro-unified internet, there are serious problems and concerns and maybe studying those things more closely, discussing about those things closely might give us a better way to think about more comprehensive, maybe the unified internet is just one or part of the package that could bring about a good social changes. Maybe I spoke long enough. Thank you.

>> AVRI DORIA: Okay, thank you.

So it actually starts to get even more a little confusing in terms of free and open is not always free and open and certainly not always good. And we certainly heard that before.

And I'm sort of starting to feel that it's starting to cover a lot and not cover much at the same time. So it's becoming actually more and more interesting conversation.

Next I have tatiana Trapina from lytton university who needs a microphone. And please, could you sort of help bring it in a little bit?

>> It's working now, thank you very much, Avri.

I was listening to everybody and thinking about this 100 flavours of fragmentation or 50 shades of fragmentation. And I think the confusion comes from the fact that we place a belief, our faith in global connectivity on different layers of the internet. And I'm taking here the democratic approach. From the perspective of technical layer nothing has challenged the global dominance of TCP/IP, the system of unique identifiers and if something has challenged it for example, the incompatibility of IPv4 and IPv6 IP addresses. This has been fixed. The technical tools have been developed for the global connectivity to win.

So for me, the glue that brings all these levels together fulfills this global connectivity is still there. But I do understand the different speakers put faith and definition of global connectivity somewhere else. Below the technical layer than internet shut downs become internet fragmentation or above layer, censorship also become internet fragmentation. Because their promise of global connectivity is somewhere else. This is where this debate gets confusing. To me internet is not fragmented exactly because the glue that keeps us together, is still interoperable and global. To remove censorship you have connectivity. But once this layer is gone, everything is gone. Yes, it's not fragmented but, but, the question is, is there no danger? There is a danger. And to me the danger is that by trying to regulate and territorializing information flows, try to exercise control be it political system or legitimate concerns about protecting the citizens from viral threats, governments start imposing restrictions that might intentionally or unintentionally impose regulation that might tackle the technical layer.

Here I have to go away from my technocratic approach and say one thing. We like to think about technical layer like unique identifiers, TCP/IP, it is all connected and glues it together, it's working. But we have to think this doesn't exist on its own. It exists not because the governments imposed it, not because regulation imposed it. It exists because the technical community, multistakeholder community put faith in it at some point. By adoption of these protocols. By adoption of this system of unique identifiers and it runs purely based on trust.

And away from my technical technocratic approach, if regulation destroys the technical underpinnings of this trust, this is why internet is going to fragment. And I do believe this is a danger here. I would like to circle back to what Elena said about self-fulfilling prophecy.

We talk about definitions a lot here. I believe at some point we start talking about solutions. And to me, one of the solutions would be to be very careful saying that internet is fragmenting or fragmented. Because it's some sort of debate. What we have to dorks we have to start thinking about basic and basic commitments. I know it's hard to fix government censoring the internet. And sometimes we have to stop label it as fragmentations sometimes it is just purely human rights abuses. It's much fancier to say fragmentation, right? We have to look in the core, the core with global connectivity and trust. And if this session can start any debate about steps forward, I would say it would be commitment by governments, by technical community, by anybody else to these basics. Once we preserve these basics we can solve any other problem because the global connectivity will prevail. Thank you.

>> AVRI DORIA: Thank you.

So we get to a point where we are really starting to over load the term and we over loaded it with all of our frustrations and unhappinesses and everything else and get in trouble.

Next I want to move to, I have a couple more before we will come back around, so many good people to talk to here, with Shitao Omar with Global Partners. Please pass the microphone that way. Thank you.

>> Shitao: Hi, everyone, sorry I'm late. I'm glad I got to catch Tatiana's input there, I think it was very helpful. The more I listen to these discussions, the more I feel like we are actually getting somewhere, as long as we are happy to navigate a choppy water.

I think one of the challenges that we are facing is that we are talking about something we are trying to preserve and also evolve. And so we are trying to figure out perhaps, as you were saying, Tatiana what we need to preserve. I think that's helpful to identify and agree on and how we evolve that considering we need to preserve that. The issue is whether through unintended or intended actions, and as you mentioned a lot of those can come from regulation, there are challenges to preserving what we have when it comes to the internet, those critical properties, the values, the principles of openness and connectivity and indeed, user control and autonomy. Those are being impacted or could be impacted by regulation and decisions and the normalization of actions like shut downs for example. So that, to me, is the challenge. How specific we are, or how broad we are, I think that comes from, yes, perhaps identifying and agreeing on what we need to preserve, identify the challenges are to that and then ensuring we can continue to evolve the internet.

According to that. So just quickly, on where I think we have come to. I co-lead the policy network on internet fragmentation and we have developed its framework, I think probably referred to before. There we have some recommendations under each of the elements of the framework which are the technical layer, we refer to the critical properties of the internet but also user experience which combines the impacts of government regulation and also corporate actions on user experience and develops recommendations based on those. And then governance as well.

The challenge of having duplicative mandates or bodies that are not inclusive and therefore don't coordinate and communicate with each other. Now if we, I believe we could take any of those and if we did some of that, that would be helpful to ensure we are both preserving and evolving the internet in a way that preserves its original vision. But it is also possible to do part of that and still go along the pathway. I think what we are trying to figure out here, what is the pathway and to have some sort of compass for that. What I hope is the policy network's contribution, which builds on the contributions of many others who have worked on this topic, world economic forum or Internet Society helps to form that compass and to both preserve and evolve what we have so we can move along the right pathway. Thanks.

>> AVRI DORIA: Thank you. I just want to mention we have two more in this initial list. Then we have an online comment. And then I want to move into a more organic conversation. And mention to the people there, there are empty seats around the table, so if you are going to want to say something, find yourself a seat. Next I have Raul Echeberria and the microphone should start moving toward him from Latino Americano internet. Raul, I tried.

>> Excellent pronunciation. That's a very interesting discussion. I say yesterday in a meeting that we have to escape from the issue of  definitions. This is where we are stuck. But we are clear about whole things we want, how we want the internet to be to behave and one of the things we don't want to happen on the internet.

One of the things we have, people have the same experiences on the internet around the world. And if it's not happening, I have experience of that, as other colleague say before, let's not put names to the countries. But I have been in countries where I have not had the same access to the same applications that I use often in my country. Or in most of the world.

And there are others, tatiana pointed out, there's risk in many policies that create impacts in the way that the Internet works.

I think we can be two years discussing this fragmentation and probably we would not get a consensus.

I will not spend time saying if internet is fragmented or not. We have problems. That's the point.

And we know what the problems are. Okay that's it. We can say, don't say the internet is fragmented because it's like to create that idea if it is already fragmented so what's the problem? But we have to be careful because, in fact, there are policies that have already been adopted in many countries that create that have huge risk on fragmentation. For not saying that we can create the opposite idea saying people complain when we pass this law and nothing happened. Everybody now is saying that internet is fragmented. So what's the problem.

So let's focus on the things we don't want to happen. We want people to have the same experience on the internet anywhere in the globe, take advantage of the powerful connectivity. We don't want interference from governments deciding by us what we can do or what we cannot do.

And also there are legitimate interest and rights of the government to take care some things proportionally we know there is a common understanding that in the world about child pornography, terrorism and other things. But also the measures that are taken to avoid access to this kind of information should be proportional and reasonable. And not use a big weapon to kill an ant.

There are problems. This is the point. We have to focus on that end up discussing if internet is fragmented or not.

>> AVRI DORIA: Thank you. So we move from quibbling over the definitions and I like the notion of combining two things. The pathway to solutions, almost. Defining the actual problems and working on solving them. The next person I have, it's the last one of this first round as it were, is Paul Wilson from APNIC, do you have a microphone? I can give you this one, or that one is coming.

>> Paul Wilson: Hi, I'm Paul Wilson from AP Nickware. I think we have spoken about fragmentation nuanced and high levels I'm not sure we want to get back to the nitty-gritty of the internet layer. I do want to say the purpose of what happens at the regional internet registries is to avoid fragmentation. Is to ensure we have an internet layer in the technical sense, the underlying layer that supports everything else that is unfragmented that can continue to grow and operate without fragmentation that can happen in various kinds. It's continuous work that needs to be done. As I tried to say in yesterday's panel it is something that shouldn't be taken for granted, it can be eroded. Looking at fragmentation as a whole and individual case, we are learning fragmentation isn't just a condition of the internet. It's a quality that varies that comes and goes. It varies by layer, by context, by geography and so on.

So that goes at the internet layer as well. And what we are continually trying to do through policy making, through the IPv6 transition, through the management of the last supplies of IPv4 is to preserve the integrity of the internet. If anybody wants to talk about specific aspects of that, IPv4 or IPv6, we can, but I feel we are past that.

I wanted to make one observation that is about the changing nature of the internet itself. And really how these things do need to be tracked and observed and analyzed as the internet grows and changes. There's been a huge trend to a kind of fragmentation of the internet over the last decade toward CDN'sment content distribution networks take copies of content and move them close to the consumers in order they can be accessed quickly. That's a type of fragmentation because it kind of breaks the model where the user it accessing a service which somehow is somewhere on the internet. And that service doesn't even more, one service on one IP address, it actually doesn't exist in one place according to the classical model. It's distributed, it's fragmented, the original end to end model is kind of fragmented by that situation. That's a huge trend and huge amount of traffic on the internet has been, these days delivered through CDN's. To the extent APNIC side Houston said we are seeing the death of transit on the internet. That is the ability of the internet to negotiate a connection from anyone point to any other point through transit networks. That's a good point, if you are no longer demanding transit end to end connectivity it may well fade. But then along came COVID. And I think the fact we had this incredible plethora of end to end, point to point video communications that became a necessity of every day life pointed out the necessity of the importance of that end-to-end to connect to any other endpoint. I was struck actually by the remote, the virtual participation by Javier before from the other side of the world on this HD connection. We have an end to end that is allowing that kind of connectivity to take place. That should be still pretty remarkable to all of us and something not to take for granted. Thank you.

>> AVRI DORIA: Thank you. You are right, it's rather miraculous and we assume it will work and get kind of flustered when it doesn't. We will move in the next. We have used about an hour. We basically had a fair number of people give a fair number of good views that have sort of boiled the ocean a little bit for us.

Adam is going to read a comment that was online. And then I would like to basically, up until now it's been just talk until you said what you wanted to say. Now with a half hour left, if you have brief points to make, vis-a-vis what other people have said and Adam will carry around a microphone. And if there are participants here who weren't the assigned speaking participants who would like to speak and have something to say, please let us know so we can get a microphone to you and speak. Either sit here or Adam will bring you a microphone. Please, Adam?

>> Adam: Thank you. Yes, I do need the exercise so please call for the mic, I would love to rush over and give it to you in a second. There is an online comment it actually covers something that Tania mentioned it's from Dhruv Dhody. He says while all of these can be called internet fragmentation, would you agree that they are not all equal and fragmentation at the technical layer that does not allow interoperability at all is a bigger threat than content moderation? Thus, is there a need for us all to be more nuanced when talking about internet fragmentation rather than sometimes clubbing them all together which does not serve us well. So that was the comment. And I think Tania touched on some of those issues.

>> AVRI DORIA: (?) kept putting her hand up. And then  --

>> Thank you, Ponsilate speaking. I just want to go back to Elena. She raised something on the other speakers really talking about it that it's not really a technical issue we know even when you try to put fragmentation, you try to package it down, you will discover most of what people actually are talking about is really human rights and digital rights being abused. My question on this political stuff, that is likely what will break the internet.

What type of political situations are you seeing. People might treat them differently. I can use a situation like what happened in Gabon, in Africa where  Ali Baba cut down the internet. He start crying use the internet to make noise to get him released.

Some people consider that impacts in Africa shutting down the internet is a fragmentation, we are depriving some people. But it's not actually that. So the political, if you can digress Elena what actual political scenarios you can see break down fragmentation, cutting off a whole continent or whole sector. Thank you.

>> AVRI DORIA: We had, Chatal, you directly addressed a question though, correct?

Yeah.

>> Kind of ended on, which is I don't think we are taking this for granted. There's a sense, even if we don't know what it is around, we are concerned we are on the wrong path. And we are moving away from something that we, maybe isn't perfect but we are moving away from it. So I think my question would be, or provocation would be, do we know, we all know we aren't taking it for granted but do we know what we need to do, do we agree on what the main issues are? And as I said, some of us have been put together recommendations including from this IGF, from the multistakeholder policy network for what can be done, is that useful or helpful to say if we implemented those recommendations things would get better? Do we have that common understanding? Because we are all clearly concerned about something.

>> AVRI DORIA: Thank you. Did you want to address something directly? I have the gentleman there, he already has a microphone.

>> Thank you, I'm from the Germany academy. I forgot the name of the lady in the black and white. You. I don't know how would you connect the discussion about fragmentation and A.I.? Because what I see is that all of the costs are being for running the internet and providing the data on all the service into the cloud is socialized and profits are privatized on very few number of companies. So we are kind of doing all the work and they are scrapping all of our data and sucking in the profits. Isn't that also a kind of fragmentation? So how would you connect these two or are we going to see these kind of connections?

>> AVRI DORIA: Okay, trying to keep track of the hands in the order I see them. I did have four, but if somebody wanted to respond to what was just asked? Yes.

>> I wanted to respond to what was asked previously, not right now.

Thank you very much for the question. I was actually taking note of what the other people were saying and trying to react to that, I was going to get to that point. But let me get from the beginning. We can keep debating how to define things, of course. By the way we do it because we were asked to do it here.

But I really Tatiana's speech let's look at what we want to preserve. The issues online brought up are very important. We have issues with that localisation, islands of secluded content, shut downs what have you, all of that. I just try to put it in perspective, in order to get to what we need to preserve. If the internet breaks into splinternets into two, three, four different internets then I think the problem becomes the whole other magnitude. The frustration we are talking about we will all be feeling will be a whole other level. We assume it just works. In this scenario we are discussing it will not just work. That's the opposite of what we need to preserve. Other people mentioned what we are discussing. Legislation to address content issues. That have an adverse effect at the technical level. I agree, that happens. I also see and no one in the technical community will say legislation is not needed. When we have that and have effect on the technical level it's usually intentional and when you explain to legislators, they fix it. The trend worries me and goes back to the question though it's been unintentional touping on the basics of the internet, the fundamentals, the identifiers we have initiatives, legislation that these targeting the identifiers. We have now in the context of the geopolitical situation if you will, we have an effort to apply sovereignty over something that is by definition global. An example is sanctions over IP as an action that goes into that direction.

Therefore that is something we need to avoid. In order to preserve what we need to preserve, going back. Thanks.

>> AVRI DORIA: Thank you. I'm starting to buildup a queue here. I have you at the microphone there. Then I have Jorge. Then Raul, then I've got Michael and then I've got Tatiana. So that's the order I've managed to build. If I didn't get it quite right, I apologize.

>> Thank you, my name is Robin green, I work with Meta and human rights issues with surveillance law access. I think this is a great discussion, thanks for hosting it. One of the things I've heard a few times over the course of this IGF and the many internet fragmentation conversations we have had is this idea that content distribution networks are fragmenting the internet. I want to push back a little on that a little bit because those networks are ultimately often times necessary to connect people to services all over the world to make sure that those services are resilient, to make sure people have access. To fast internet service. At the end of the day, when we are talking about internet fragmentation, in my view one of the things we are focused on, when talking about regulatory implication a core technical fragmentation of the internet like some folks have talked about, the thing we actually care about is what is the user experience. Is the user experiencing fragmentation and if the goal of the internet, which in my mind is for people to exercise their fundamental rights and whether those are economic rights, expressive rights, you know, accessing information, engaging in assembly and things like that, at the end of the day, if their user experience is becoming fragmented in a way they can't fulfill those goals then to me that is internet fragmentation that needs to be addressed. So we can sort of have this larger umbrella of internet fragmentation while still looking at things from a technical perspective and user experience perspective. But I think it would be a mistake to step away from the concept of internet fragmentation because something isn't directly mandating a technical fragmentation even where the user experience still winds up being fragmented. That's where I see data localisation requirements or other kinds of restrictions on cross border data flows, restrictions on encryption that would be implicating people globally, users globally and similarly, implications on content take downs and geo blocking and other restrictions of free expression. Those are all elements of internet fragmentation, whether technical or user experience oriented, there's a difference there. But there are still things I think we all need to consider.

>> AVRI DORIA: Thank you. Next I have Jorge, I was reminded there's a pending question in the air, if any of you have it in the connection of A.I. which has been a favourite subject. You will have one, fantastic. We will get there. I was reminded that I had not made sure  --

Okay, so we will get there. You want to do it now? Okay.

>> So it doesn't get lost. Because it was asked. I really doubt how would you connect the discussions on fragments in A.I., I would say I would not connect them, sorry. That's my answer.

But I'm still in the queue for the other issues.

>> AVRI DORIA: Thank you. Jorge, please.

>> Okay, after this commentary. Jorge with content, so many things have been said, it's difficult to add something. But I couldn't resist. I think here, as we are in the U.N. in the IGF we have to connect this also to the discussions we are having at the global level. What's happening at the global level. And if you look at how the situation is evolving, just four years ago we had a report from the high level panel on digital cooperation which was called the age of digital interdependence. So really, the focus was made laid on what unites us on how dependent we are with each other and through the digital issue that unites many things.

And I just wanted to mention that because the situation today is a completely different one. Even if such a panel would try to name its report the same way it would be criticized as being completely out of the reality where we live with very fundamental geopolitical tensions. I just want to share that and recall we are in the midst of this process toward a global digital compact where digital fragmentation is one of the topics to be considered. And going back to something that Paul said before, and others. The internet interoperability at the technical level is not freely something we should take for granted. It really relies apart from this history of trust of building this network. It reare lies on huge network effects. On incentives and benefits for everyone  connecting to this unique network. But really the pressures are mounting at this geopolitical level. So there may come a time where those pressures, perhaps joined by alternatives at a standards level, at other levels become so important that this delicate fabric of trust, which holds the tissues together are built by millions of networks begin to erode. So this is something that I think is the fundamental level of internet fragmentation. The same way this is fabric of millions and millions of networks, it is also in the hands of those millions of people with networks with companies with their governments who can take decisions going into the right direction or into the wrong directions. And can decide to invest into holding that tissue together or continue eroding that tissue into a direction that may end up with a fragmentation.

So perhaps this decision or recommendation of investing into the right direction which is something in the hands of many of the people coming here could be something for the policy network on internet fragmentation and for some good recommendations, useful recommendations coming out of this IGF and flowing into the GDC. So hope that was helpful after so many thoughtful inputs.

>> AVRI DORIA: Thank you. Starting to have a very long list here and a very short amount of time. I've got Raul next.

>> Raul Echeberria: The colleague who raised the issue left the meeting but I wanted to come back to that point. Of course in countries that are not democratic  --  it's important to impose restrictions to access to content. Something by the way we naturalize and we should not. It's not the only problem. This is something that we expect. But now we are facing problems in strong democratic countries that are passing laws and developing policies that are really affecting the user experiences. Sometimes it's based on measures that try to protect intellectual property in the networks or because some taxation or other things. But because the effect the policies could have on the internet though things are adopted.

I'm sorry to say my experience on policy makers is not only as successful as the colleague from ICANN say many times was explained to policy makers and we are not  --  they have, as I say yesterday in another meeting, the incentives of policy makers are diverse. Sometime thez have political decisions to protect industry or protect from disruption, so they have commitments. They have to move ahead with the decision even knowing they are creating an impact.

As a friend of mind usually says, sometimes when policies don't feed with reality, some policy makers will try to do, try to change the reality instead of changing the policies.

So it's a proob lem, I want to come back to the political situation. Not only a problem with dictatorships or totalitarian rejet streams having this fermented experience. Nobody care about the bites, we care about what people do on the internet. This is what really matters.

>> AVRI DORIA: Thank you. Next I have Michael H.

>> I'm from the association of the internet industry in Germany. I'm working with the internet since 1983 and when we started the only fragments, there was no internet, it consists only of fragments in the various countries, fragments of networks, everything. What we did at that time was building gateways. Of course that may not be efficient. There is a risk of filtering, I admit but I'm pretty much sure fragmentation at the technical level goes on, someone will find a technical solution for that one. And then we only have to deal with the political stuff. Thank you.

>> AVRI DORIA: Thank you. I think I come from that same generation, for me the internet is constantly becoming. Tatiana?

>> Tatiana: Unfortunately the gentleman who asked the question for feasible scenario left. But I wanted to say we can think about technical scenario, alternative standards, alternative system of unique identifiers. I don't believe speaking exactly. I think the technical community has enough experience of connecting things by coming up with technical solutions so the connectivity wins and plus, if this is imposed by the governments, it will incur huge costs. I do not think this is completely unrealistic, maybe in the future, if one region like the European Union decides to go with absolutely technical standards it might happen. Where I see the physical scenario, this is where I think it becomes very important is regulation is imposed, targets technical layer in a way that would call for breach of trust is eroded. When something on the technical layer (?) unique identifiers, IP addresses in certain regions have different frameworks for governments, when the multistakeholder governments doesn't cover it all or competing frameworks with what we have now. Here it brings me to the point what we want to preserve. I think we want to preserve essentially what makes the internet the internet.

--  I think much more important in terms of feasibility of any, I reluctantly say this word fragmentation scenario. We need to preserve this trust. We need the firm commitment to the multistakeholder model of governments, not of engagement or discussion but of governance because this is how the technical layer has been governed and this is what we have to constantly recommit ourselves to. Thank you.

>> AVRI DORIA: Thank you. As we come closer and closer to the end with seven minutes, please Tomaki.

>> Thank you. I wanted to answer two questions. One, the splinternet discussion, how it relates to A.I., I think I have slightly different take. And there are two relations I can think of, number one, A.I. especially the current like large language model kind of A.I.'s, they are built up on a massive training dataset. Which is enabled in a way by the unique internet. If we wanted to leave some benefits from it, from those A.I.'s it's important that internet is large and as interconnected as they are right now. And also more to the political domain. I think it's good that some of the A.I.'s can provide at least some advanced capacity to translate and overcome language barriers, which in a content layer level, I would say, connects the internet even more. So that's my take on the relationship. And the other question I wanted to address was if the technical layer fragmentation matters more than the content layer, and I think the answer basically is yes, but not in a simple way. My understanding is that if, suppose if government have granular and speedy capability to regulate online communications of any kind, then the government doesn't really care about shutting down the internet connections because such a major is always a blunt instrument compared to very granular control of the communication. But of course no government has such majors so it matters that internet is connected to each other. Thank you.

>> AVRI DORIA: Thank you. Okay, I think I have four people left. The next one is Javier online. Then I've got Paul, then Dushon then Uta, if I get all those in we are really doing great. That means you are speaking briefly. Javier are you ready?

>> Javier: Yes, I hope you can hear me, I would like to add in the side of solutions just to kind of move into something different. When it comes to what we have identified as the core, fragmentation threats, in terms of what happens with the protocols and identifiers, we have heard that the main threats on that front come from the governments, right? Governments feeling sometimes impotent when it comes to try to control the internet in order to control public priorities. So maybe isn't then a reinvigoration of the multistakeholder model. More attention to that, active Association of those enactments by governments, extreme advances by governments and also getting more information to users or those who can exert pressure on their own governments a way of offering solution, we should be doing that. Just thinking about what is the main threat to more specific technical aspect of this. And maybe with more participation by governments we can make a valuable contribution to that. Thanks.

>> AVRI DORIA: Thank you. Paul, please?

>> Paul: CDNs are useful and I didn't mean to indicate they are not but they don't help people access internet.

>> If we are talking about user experience, fragmentation is a lack of interoperation, single instant messaging account like an email account and still exchange messages with those that choose different services whether WhatsApp or signal or anything else. Those could interop rate and they don't, they generally don't and that's the choice of company's concern. And I think that will continue and that will continue to me to represent a fragmentation of the experience on the internet. Those companies will continue to do that until they are required to change that behaviour. And I wouldn't mind seeing that.

>> AVRI DORIA: Please.

Duchon (?)

>> Dushon for the record from Serbia. I would like just to express one frustration that I have about fragmentation. So we call everything fragmentation.

We call filtering fragmentation. We call, I remember and I agree previous talks that whenever, when I was involved in internet, it was fragmented. Later on we were talking about Balkanization, if you remember, 2014-2015 and 2016. Domain names for example are still fragmenting the internet. So technical layer is still the protected layer as I would say. And it is connecting everything but we have given governments to legislate the part of the internet, so that part of the internet can be and should be fragmented. On the other side we are fragmented with filtering, don't call it fragmentation. Blocking or something like that. Let's talk about those particular topics. Not call it fragmentation. We will have high level discussion on everything without substance.

>> AVRI DORIA: Thank you. Uta, you get the next to the last word because mine will be last.

>> Umai: Since we have been collecting indicators for fragmentation been writing up research about this, I would like to put one more point next to this, we have been focusing on the technical layers and while that of course is very important, I find it important to mention, the social layer underlying the internet as a network of networks and that consists of network engineers who maintain these systems and have a huge informal community with informal values and forms of coordination that may be aging, so if we are looking at this in the future then we may be wanting to look at this community as well and their capabilities of actually keeping things together.

>> AVRI DORIA: Thank you very much. And thank you all for a great conversation. And I'm certainly not going to sum it up because that would take forever. But the simple mind of mine walks away fragmentation is a four letter word with lots of nuance and lots of views. So thank you very much.