IGF 2023 – Day 2 – Policy Network on Internet Fragmentation

The following are the outputs of the captioning taken during an IGF intervention. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid, but should not be treated as an authoritative record.



>> BRUNA MARTINS DOS SANTOS:  This is the PNIF.  My name is Bruna Marlins dos Santos.  I am one of the co‑facilitators, and we would like to welcome everybody to another discussion, it's a very happy moment for us.  It's the second year of the policy network, and we are very glad to be presenting today a little discussion on how the framework has evolved and to do a defend dive on the discussion paper that we, the policy network just put out.

So the session today will be a little bit of a presentation of the discussion paper and also some debates between both the pen holders of the document and some community commentators on the three aspects we describe in the policy network framework.

And before we move onto that I wanted to start with a very big thanks to every single volunteer of the policy network that helped us shape this document.  Some of them are on the stage with us, some of them are here in this room.  Thanks a lot for joining the conversation and helping us construct this debate.

I'm going to hand the floor to you, Wim Degezelle, right, and we can move on.

>> WIM DEGEZELLE: Thank you.  As you see we have a presentation, I think the first, Bruna, thank you, you gave the overview of the agenda.  So I will give the brief introduction.  My name is Wim Degezelle, I'm part of the policy network, their intersessional activity by the IGF.  That means that they also receive support by the IGF Secretariat and I'm with the Secretariat as consultant to help this policy network to start.

So a brief introduction on the policy network on Internet fragmentation.  It is an intersessional activity.  That means we not only work at this IGF meeting, but we start way earlier.  We started working in May, and even before that to prepare and work to this session and to the IGF.

So it is the policy Network Internet Fragmentation.  There are other policy networks also on the agenda, but this one on the Internet fragmentation has the, or wants to further the discussion and raise awareness on fragmentation on technical policy, legal and regulatory measures that may and attacks that may pose a risk to the open to the open, interconnected, interoperable Internet.

So what are the objectives of the policy network?  As Bruna mentioned this is the second year of the policy network.  What we wanted to do was to understand what is actually meant with Internet fragmentation so come up with a comprehensive framework and overview of what Internet fragmentation is.  We look at case studies, what actually is happening and try to come up with examples or look for examples, and then the third question is what to do about it, how to address or how to try to avoid fragmentation.

This looking back to what we did last year, we actually dove into those questions, and as often is the case, you want to find the definition and try to define what you are working on.  Through the webinars, like you see we had webinars during the year asking specifically the question of what does it actually mean to people when they talk about Internet fragmentation, and what should, and what can be done about it, and who should be doing what?

Very quickly through those webinars and those discussions we had, it became clear that trying to come up with a definition is not really something that is still possible.  It might have been possible earlier on, but how the people are discussing and looking at fragmentation, trying to squeeze that all in one clear definition is not helpful.

What we did instead or through the work it became clear that there are different views on what fragmentation is, and that's how last year, as the outcome of last year's discussion, we came up with a framework saying actually, if we listen to the people, if we listen to the comments we get, they kind of can form baskets of what people see and understand as fragmentation.

What's in those baskets we will further discuss and hear from the panelists today.  But, and I think that was the main output of our work last year, that allowed us to come up with a framework.  The framework you see in small, and this is a larger version.

So the framework that said, well, we found that when people are talking about fragmentation we can either form a basket that we can label as fragmentation of the Internet user experience, fragmentation of Internet Governance and coordination, or people really refer to fragmentation of the technical layer, technical architecture of the Internet.

Those were the baskets we could form.  With the important comment we got, those baskets are not alone, are not completely separate.  There are interactions, there are overlaps between them, and that shouldn't be considered as separate silos.

One comment before I hand over to Sheetal to discuss what we actually did this year, is we labeled the framework as it is a framework for discussing Internet fragmentation.  We don't want to come up with a framework to define what it is, but we from the beginning said, well, this framework should help to discuss and further the discussion.

Because I think that's one of the main evolutions we saw in the work in our discussions is that people started to move from, well, we need to define something, and then we need to discuss and kind of an understanding that it is important to discuss with stakeholders, and have these multi‑stakeholder discussions on Internet fragmentation, but these stakeholders are not necessarily always the same.

It is possible in those three layers of our framework that you need to sit together with different stakeholders or different types of people, different organisations.  And I think that's one of the major or one of the main findings we had last year in our work.  Together with, and that probably will come leer out of today's discussion, the second point is what those different groups or in those different layers they come up with actually guidance or guidelines or ideas on how to avoid or address fragmentation does not necessarily is completely complementary with each other.  So at the end of the discussion, it will still be necessary to cross those baskets have discussions on how to address.

So that's what we did last year.  I hope that was clear, so this was the framework and that was the start of our discussions this year.  So I hand over to you, Sheetal.

>> SHEETAL KUMAR: Hi, everyone.  Thanks, Wim.  It's great to be here, and to be presenting our output for this year, and as Bruna said, we co‑facilitate this policy network, and it's really nice, I think, to now not be doing so much of the work, but to be hearing from you.  Once we've heard from the drafters and the commentators who will be responding to the drafters of this year's output, we really want to hear from you.

The work is going to be in the room, and then as you can see from the agenda as well we will also be looking for feedback after this session.  So what have we done this year?  As Wim said, we have been building on the work of last year where we developed a framework to conceptualize what Internet fragmentation is understood as by, as we have just discussed in many different ways, and so this framework we developed is to support, it's really a tool to support better understanding and clarification of what Internet fragmentation is.

And in that sense, what we were able to do this year is further unpack what the framework is, and those three areas which Wim outlined, and those are the fragmentation of the technical layer the user experience and governance and coordination.

What we wanted to do in unpacking these areas was better understand what the priorities should be in each area.  So what is actually harmful and negative, and from that assess what can be done, so develop some recommendations for action, and where we are really, I think, looking forward to hearing from you all and from those who have been so involved already is, well, really whether or not you think that these recommendations are helpful, whether anything is missing, and whether you think the way that the different elements of the framework have been unpacked and what has been prioritized aligns with your view of what we should be focusing on as an international community when it comes to this issue.

So what we are going to do is take each, each element of the framework one by one and I also invite you to go to the PNIF web page and look at the discussion paper as we are discussing it here, and consider also in the second part of this session how you may want to react to what is being presented.  So we are going to do, first of all, a presentation of each track or each element.

So we have the very hard working drafters of the output document here, and we are going to go one by one, hear from them.  They are going to just present the top level findings or the top level points, so what priorities they found they need to be addressed and then some of the recommendations.

And then we will have a commentator to respond, and so we will do that for each, and then we will open up.  So without further ado, I'd like to hand over to Rosalind Kenny Birch who is with the United Kingdom Government at the department for science, innovation and technology.  And, Ros, you worked with others to develop the chapter in our document focused on Internet Governance and coordination, and the fragmentation of that.

So in the next three or four minutes would you be able to provide an overview of what that chapter says and the recommendations that you have for addressing this element of fragmentation?  Thank you.

>> ROSALIND KENNY BIRCH: Thanks Sheetal and great to see everyone here today.  One of the points of this panel discussion is to evoke a conversation.  Our multi‑stakeholder Working Group that worked on this chapter had quite a few different perspectives, because fragmentation is such a complex topic to discuss.  So it will be really interesting to hear your insights here today from a wider group of perspectives, and so I would really invite you to engage in the discussion, offer some of your own insights and challenge afterward as well.

But just to present on what we've written up in the preliminary draft chapter on fragmentation at the governance layer of the Internet, I would first like to lay out a little bit of context.  So our multi‑stakeholder Working Group, draft wrote that fragmentation at the governance layer primarily relates to the interactions between global Internet Governance and standards bodies.

When these bodies do not coordinate inclusively, it can and does result in fragmentation.  This fragmentation can manifest in siloed or duplicative discussions, and exclusion of specific groups from participation resulting in decisions being taken without consensus from the multi‑stakeholder community.  And it's important to note too that fragmentation at the governance layer can also create knock on effects at the other areas of the Internet, user experience and the technical layer.

So there were a couple of different components to our analysis about fragmentation can sort of emerge or come from at the governance layer.  One was duplicative mandates.  So if part of specific Internet Governance body's mandate is unclear or may have overlapping elements with a different body, this can foster a competition for legitimacy or create confusion between bodies.  And, therefore, that can make it difficult for stakeholders to know where and when they need to engage in a specific conversation.

Another point we observed was when mandates are exclusive or don't fully empower all elements of the multi‑stakeholder community to participate.  So we see inclusion as central to combating that so that people can participate on an equal footing.  And then finally, taking actions at the right level.  So individual Government actions can sometimes lead to divergence in the rules applied to the Internet and its management, and in that sense, it's really important that national Governments and Internet Governance global bodies are closely conversing about issues specifically when they are being developed or discussed through multi‑stakeholder processes already.

With some of the analytical points we proposed a couple of recommendations, and, again, very eager to get perspectives and feedback on this today.  One was not to introduce further bodies into the Internet Governance landscape.  The Internet Governance landscape is already complex, and as we all well know through all of our travels, there are a lot of different conferences, events taking place that we engage with across bodies already. 

And people only have so much time and only so much financial resource to be able to engage in these.  So further perpetuating that complex landscape could end up excluding people from discussions if they don't have the resources to fully participate in more and more emerging bodies and spaces.

However, that being said, another recommendation we made was, therefore, it is important to improve coordination between existing Internet Governance bodies to help address perceived or real gaps in these spaces.  So coordination between existing Internet Governance bodies is needed to help address that as well.  Thirdly, and in order to avoid siloed public policy discussions regarding Internet Governance, all Internet Governance bodies must be fully inclusive to stakeholders, and enable meaningful multi‑stakeholder participation on an equal footing.

We also believed that that would help address instances of fragmentation at the governance layer.  Then finally we recommend that existing global Internet Governance bodies must engage more closely with national Governments.

So this goes back to our point of analysis before.  There is actually a two‑way street here.  National Governments when looking at proposed legislation can benefit from talking to global Internet Governance bodies about their plans and, therefore, receive important information and feedback.

But equally, global Internet Governance bodies should be on the front foot about engaging with Governments and ensure that Governments know what activities are going on in the global space to hope avoid duplicative measures.  So I will stop there and an exciting part of this panel is we will receive challenge and other perspectives on this work.  So with that I hand over to Jordan Carter, great to have you here.

>> JORDAN CARTER: Thank you.  Good morning, everyone.  My name is Jordan Carter.  I work for the AU domain administration, the ccTLD manager for dot AU.  It's a pleasure to offer  provocative provocations to the group to help the conversation happen.  I am making personal remarks.

Overall, I think this is a good start to the discussion around fragmentation, and my congratulations to the volunteers.  I should disclose aside from joining the email list a couple of months ago, I have not been involved in the paper.  I was reading it fresh to prepare for this session.  I agree with the analysis so far as it goes.

So in the end, my provocation is relatively brief.  The importance of broadband participation is vital particularly in the standards bodies and in some of the global Internet Governance organisations like ICANN, the western bias in participation is undeniable, and meaningful participation from around the globe and from groups not participating is absolutely essential within whatever framework that we have.

When I read the very first box, the definition fragmentation of Internet Governance primarily rates to interactions between global Internet Governance and standards bodies.  My core thesis might be that that's too narrow a definition of governance fragmentation.  Because one of the key agents of governance are Governments, and to not deal with Government‑driven policy, driven fragmentation in this session I think maybe complicates the picture and I'm sure I can in turn be challenged about that.

And part of the challenge there is that the definition of Internet Governance itself is under challenge.  Do we think that it's just about the governance of the Internet, which is a distinction that's been made, or is it the governance on the Internet or is it these broader questions of digital governance that get often tacked onto those infrastructure level discussions today?

Another challenge I think it would be worth taking into account in the governance fragmentation is that caused by the narrow mandates of a lot of the technical Internet Governance organisations though marrow mandates are there for good reasons, but sometimes they make it difficult for those organisations to actually deal with the systematic view of what's going on in the Internet.

So you can have a situation where each silo is dealing with its narrow mandate and none of them are prepared to take a view about the system as a whole.  And so I think there is some institutional drivers there at the global Internet Governance level towards fragmentation.

The taper talks about the need for better coordination, and I agree.  It suggests further research, and I agree.  But quite a lot of the people who are involved in these global Internet Governance bodies could undertake meaningful card nation together without further research.  They just need to start doing it.

Some of it is being done, but the challenge not to this paper but to those organisations is get coordinating in the face of the challenges that the Internet is throwing up, and in the challenges to the governance model that we see today.

And I really did appreciate the paper calling out the duplication and risks with some of the proposals in the Secretary‑General's policy brief for digital cooperation forum, for example.  The last thing that we need is duplicative institutions being established with new resources going to fund them instead of the resources that the IGF, for example, is crying out for and could make good use of.

And the last point I want to make, I guess, and having argued that the governance discussion could maybe a broader look is the multi‑stakeholder‑driven Internet Governance system and the multilateral state‑based regulatory and legal system, I think need to be much better at working effectively together.  The two can and should shape each other.

The multi‑stakeholder dialogues and organisations like the IGF could usefully inform policy if more of the people doing public policy relating to the Internet were aware of their work.  I will wrap it up.  I don't know if that was provocative enough but thank you for the chance to comment.

>> SHEETAL KUMAR: Thank you so much, Jordan and we will be going through each of the elements of the framework first before opening up.  I wanted to let you know that we had some written feedback from the community when we published the paper and wanted to weave in some of that into this discussion as well.  So there was one point of feedback relevant to the Internet Governance and coordination chapter.

It was really about providing concrete examples of hue governance fragmentation causes Internet fragmentation and just it was checking that what our understanding in the paper that you put out in the paper is of Internet Governance and coordination fragmentation is essentially that big uncoordinated national processes is a part of fragmentation.  If so why that treated differently than Government and corporate sourced fragmentation which are addressed under user experience which we will come on to.  So there is a question about what is the focus of this chapter?  Is it on the existence of multiple uncoordinated processes, which I think you have addressed, and that is the focus.

And then Jordan, you mentioned the importance of ensuring coherence or at least engagement and coordination and it might be interesting to hear from you later, but also from everyone here and online whether you have any ideas for concrete mechanisms or examples that are already existing for how that coordination can effectively take place.

So without further ado, we will move on then before we open up to the second chapter, and we have here Vittorio Bertola who was one of the co‑drafters in the group.  And I know Vittorio wears many hats so I don't know how you prefer to be introduced.  But please do provide your, well, please do choose your hat.

And then an overview of the work you've done in this year to assess the priorities and the user experience on fragmentation that we had outlined last year, and then also the recommendations that you put forward.  It's a very hefty chapter of the discussion document.  So good luck with summarizing it in three or four minutes.

>> VITTORIO BERTOLA: It's pretty hard, but maybe my head is like having gray hair from too many years in these discussions, almost 25 now.  But I work for Open Exchange which is a German open source software company.  I was one of the people that tried to tackle this problem of user experience fragmentation, which I think is the hardest and most vague one because the entire discussion on fragmentation started from the technical level and then multi‑stakeholders tried to add more things into it, and user experience things are mostly coming from this kind of approach.

So we tried to go for a different completely open and broad by saying that anything that makes to different users of the Internet see different things when they try to access the same service, website, or do the same thing on the Internet is a form of user level fragmentation.  If you take this very broad approach there is a need to tell between the positive and negative cases. 

There are many situations in which this difference in experience is a good thing.  It's made to help a user to customize content for them or made to protect the user to give them rights, for example, through privacy laws in specific countries or it's done to prevent them, to prevent them from accessing unmoral websites, whatever.

There could be another definition that finding a definition that covers only negative cases but we found this becomes harder and harder.  We would rather take a case by case approach.  So by starting from this very broad definition, we identified several priorities and different cases and then we want to work on them one by one because they all have a different need and different view to be taken into account.

So we identified two major sources of this kind of fragmentation, and it's never used.  Usually it's either a Government that for some reason wants to exert sovereignty and mode few the experience for their own citizens or it's the company, usually global platforms that wants to be this ecosystem that basically prevents user from going somewhere else because they want, of course, to exploit them for business reasons.

So through these two opposite pushes, a number of phenomenons emerge.  We identified six priorities and the three top ones, the ones we would start with are first of all Internet shutdowns.  We discussed a bit whether it's a user experience level thing or a technical thing.  If it was about India we decided we could discuss it at this level and we think they are a negative thing.

So we received a comment from someone in the community saying there is something like a positive Internet shutdown.  I don't know what it is, but with will be up for discussion.  Then the second priority where I identified the case in which national blocking or law enforcement orders have global effects so spilling over to other jurisdictions and creating issues of, I mean, problems to other countries and other citizens, and then the third case was the world governance I mentioned, so basically the building of barriers and the restriction of user choice and competition, both by Governments when they have like laws, the global ones but also by a global Internet platforms.

There is more because we also would like to discuss national level censorship when content gets blocked for political reasons.  We would like to see the regulations on Net Neutrality, which is other issues and the other is geo blocking for intellectual property resource.  So there is a long list, and I encourage people in the community to participate on specific issues.

We don't think we can make suggestions for everything at the same time but we tried to identify five principles.  Basically the idea we would like to start with is there should be principle of equality meaning default that everybody should be able to access everything in the same way.

The second one is principle of enhancement, so when the differentiation, customization is done in the interest of user or asked for by the user, then it's a good thing, so we don't need to worry about it.  The problem is when this gets imposed onto users by third party against their wish and this case could have negative effects.

So the third suggestion is we have an Impact Assessment when you do something that deviates what a national law or business decision, then there should be harmonization.  So in regulatory terms, as far as possible on global agreements on how to take care of the same problem in the same way everywhere and only go to national regulation when harmonization is missing or doesn't take into account international needs.

The most important principle is in the end through should be free choice.

The users should be free to choose how they use the Internet and where they go, unless there are important reasons to prevent that from happening.  In the end, the user should be trusted and be able to do the good thing.  Thank you.  I think we have a cement Tater, and we will give you the floor.

>> MARIELZA OLIVEIRA: Let me start by saying my name is Marielza Oliveira.  I am in the communications and information sector of UNESCO.  This work for the Public Network Internet Fragmentation is particularly important to us because what my team and I do is essentially defend freedom of expression, access to information and privacy, and these are the rights that are most directly impacted by fragmentation.

So first, I want to say also a big congrats to Bruna Marlins dos Santos, Sheetal Kumar and Wim Degezelle who have been steering this for the last year.  Let me say that to me the user experience fragmentation is maybe the most interesting type just because it has this positive side when users are served with custom features and content, and the negative side when users prevented from accessing certain features, and service and content.

And the discussion paper is actually concerned very much primarily with the negative side, which is essentially about how these features, these mechanisms actually impose barriers that isolate or trap users into an information environment from which they can't really escape.

And a consequence of isolation, and a major source of the harms that happen as a consequence of this type of fragmentation is essentially that it enables serving trapped users different world views than are served from other Internet users.

That brings a really important point that maybe it's not quite explicit in the paper yet, but I like that it was mentioned in the presentation just made is that negative user experience fragmentation actually affects all users, not just the ones immediately deprived of access to the Internet or to a specific content and services.

Some of the users that are included are prevented from enjoying their human rights to access to information or the freedom of expression and other rights, and they may end up being driven to echo chambers and elements like that.  It's also true that the non‑targeted user whose are deprived of their rights to freely associate with those who are isolated to seek information from them and impart information to them, and, therefore, the consequences that these two groups end up driven apart.

There is an increasing gap in the information and knowledge between them.  That separates people, and many times, especially when done for political purposes, the likely consequence is polarization which spills beyond the Internet and into the real world and actually may affect even non‑internet users.

I think this is a particularly important topic.  In UNESCO we work with what we call the Rome principles for Internet development in which the Internet should be human rights based open to all, accessible by all, stakeholder led, and the user experience fragmentation is very much about the explicit decisions that reduce openness and accessibility which then has consequences for human rights and when we talk about fragmentation is essentially not done by a multi‑stakeholder process.

It tends to be a unilateral decision process.  One of the things that I really like about the paper is that it actually laid out principles specifically for fragmentation that were mentioned in the presentation, in particular this issue of free choice and quality of access and enhancement of experience and others.

These are very much in line with the existing principles and particularly with the human rights framework.  And the paper actually received a number of comments already including a suggestion that this regarding user experience be explicitly based in human rights standards and processes which are already globally accepted and there is like a solid jurisprudence foundation around them.

And in particular it just said that we need to consider the three‑part test on legitimacy of interferences with freedom of expression.  So this is an element that I think it would be important to add to the paper as well.  Some of the points that have been already made through comments is that there is some content that actually is relevant to block, that is legitimate to block because there is a law that prescribes their blocking.  They pursue legitimate aim and they are in line with a democratic society.

And content like that, for example, has to do with child pornography, terrorism, incitement of violence and things like that.  This has not been reflected in the paper on how we are going to be disambiguating between those two types and the next type in the next draft should be including some of that, and maybe even making reference to the speech and debates around what is unlawful versus what is lawful.

And maybe just to finalize I think the paper would benefits from bringing up some of the potential mitigation measures, colleague, for example, talking about enforcing platform interoperability, data portability, strengthening users, information literacy that can counter act effects of the oak could chambers and disinformation and other that are created by fragmentation.

So I'm going to end up here because I know that we would love to hear the comments from participants as well.  Thank you very much for the chance to comment.

>> SHEETAL KUMAR: Thank you, that was great.  And you were very positive about the chapter, and I think you also helpfully reacted though to some of the feedback we got online, the written feedback which I have to say was really helpful and constructive.

So you can also access it on the Web page.  Quite a lot of it focused on the need to be more explicit about the use of different terms, the connection between the human rights standards and negative user experience fragmentation, and explaining the difference between what is called the negative and harmful fragmentation in terms of user experience and as I said, being more explicit about that.

So it was great to hear you respond to that as well because I think when we come to you on the floor and online, please do sort of pick up some of those points or add your own, but certainly a lot of really helpful feedback already from you, so thank you for that.  So we are going to move now to the chapter that looked at technical layer of fragmentation and Olaf Kolkman is here with us to present the chapter and really looking forward to hearing from you, and then you are going to be joined afterward or we are going to be joined by Suresh Krishnan from the Internet Architecture Board who will respond.

Then we will open up, so please do get ready with your reflections and questions.  Without further ado, over to you.

>> OLAF KOLKMAN: Thank you very much.  My name is Olaf Kolkman.  I work with the Internet Society, I'm a principal there on technical infrastructure.  When we speak about technical infrastructure of the Internet, that is the network of networks that are internetworking to provide global connectivity.  80,000 networks that interconnect to provide global connectivity, and the supporting infrastructure that makes that happen.

That's for us the sort of Internet technical infrastructure.  Now, a few ideas that we had in constructing this chapter, I want to highlight those without going to the details of the chapter itself.  But first, I want to urge people to review this.  This is a work in progress, and it becomes stronger when stakeholders engage in the document, with the document and provide comments.  At this moment I feel that I have been too little eyes on this chapter, and we can use help.

Anyway, the document or the chapter starts with saying that the technical fragmentation is not something that is clearly defined.  There is an operationalized definition of fragmentation around.  It's a work by Balter and Hideman, but they have a criterion that says if 50% of public IP addresses cannot reach the other 50%, then you have fragmented Internet.  That is very, very fragmented Internet.  That means that half of the population cannot reach the other half of the population.

And I think we don't want to be there.  It's like you are losing your hair and at some point you are bald and that 50% point, that's true baldness, I would say.  So how to prevent getting bald.  That's sort of the question.

What we also said is fragmentation is not necessarily everything where people choose to not interoperate, and not internetwork.  There are cases like that.  Like my home network, my Home Automation Network does not need to be on the Internet directly.

That's a choice.  That's a choice you can make.  Yesterday in a sense on fragmentation somebody said you have good fragmentation and bad fragmentation.  I sort of like that idea.  Decentralization is not fragmentation.  Lack of connectivity because you choose not to connect is not fragmentation.  Temporarily having to reroute your traffic because of the network problem so to speak, not fragmentation.

But what is fragmentation?  How do we define it then?  Well, again, that's very difficult, but we, the approach that we took is using the critical properties as one of the frameworks.  There are multiple frameworks we point to the critical properties that the framework that the Internet Society developed that basically defines critical properties of the Internet in non‑technical terms.

They are inspired by the network architecture, and I will not go into the details of them.  That's one of the frameworks where you can say if you lose these critical properties, if you are sliding down the scale away from these properties, then you run into the risk of fragmentation.

So this is the approach that we took.  Another framework you can look at, and approach is that of the public core.  The public core is a framework developed by a think tank in the Netherlands and later further analyzed and defined by the Global Commission on the Stability of Cyberspace.

That's another framework and lens through which you can look at the Internet and say, okay, we are impacting elements of the public core, and that might lead to fragmentation.  I think one of the things that we have done in this document is also, by doing that, by using this type of non‑technical frameworks, frameworks that do not specify actually the technology that's being used, we allow for evolution.

The Internet really is still evolving.  I think that's important that we don't ossify as we usually say the Internet in its current state.  We need to continuously be able to evolve it.  Another aspect of fragmentation that we looked at was basically what I would call the evolution of the edge whereby what we see is that there is a lot of changing in routing behavior, private companies building transit or building their own network compared to using transit to get close to the user.

And that might cause a fragmentation of a different sort with basically the digital divide, increasing the digital divide of users that have, are close to that type of infrastructure and users that are not.  And that has impact on the application layer.  There might be users that have a very good user experience, and there might be users that do not have a good user experience.  And that is true by the way that the Internet evolves in more richer parts of the world or versus less connected parts of the world.

Hard to catch within those critical frameworks that I just mentioned, but it is a point that we point out in the document.  Going to the recommendations, so the recommendations are basically look at these frameworks, use those frameworks, these critical properties or the public core and make sure that together we protect these properties.  Make sure that we can continue to internetwork and provide a global network to everybody that brings the opportunities to actually do all of this user stuff.

If we fragment only user layer, but still have a global network that connects us all, we have a chance to defragment on that user level.

But once we have fragmented, the Internet technical infrastructure, that fragmentation will also be reflected in the user space.  So it's much more ‑‑ it's not much more ‑‑ it's very important to take care that those properties are protected and we have to do that together.

There are very few ways to understand how the fragmentation is happening.  There are very few measurements around that look into longitudinal scale on what the evolution is that impacts, that fragmentation, impact fragmentation and how it evolves.  This is really a call for people to set up measurements and think creatively on how you would assess these fragmentation on a technical layer.

Once proposals are introduced either on the policy or on the technical layer in standardization efforts, for instance, do assess them against these critical properties, do assess them against frameworks and see if we lose interoperability.  See if we lose the ability to connect.  If that is the case, perhaps it's not such a good idea.

And, of course, we are into this together, and the multi‑stakeholder approach is a good thing in order to make sure that what is being delivered both by the private sector developing these technology and the technical communities working on technologies as well as civil society and Governments to make sure that we stay globally connected and don't split up this network of networks.

I think that's it.

>> SHEETAL KUMAR: That's great.  We have Suresh Krishnan online.

>> SURESH KRISHNAN: I am here.  Thank you.  Thanks a lot.  Thanks for the excellent summary.  There is very little fault in this so I will go over a few things that I think are important and then give some minor commentary.  The Internet is a live set of networks.  There is no set point of choke control over this.  This is multiple people that put together and build the large network.  I think that's a key thing to protect.  That does not mean fragmentation.

That's by design, these networks are like independent and decentralized and what holds them together is the technology that upholds the interoperability.  That's something that we got well done in the first piece of this where we talking about technology being the thing that holds stuff together and noted administration of it.

And I think that's a key point to emphasize.  The second thing is on the critical properties of the Internet, I think openness is one of them, and also the incremental stuff.  Enticing to your lack of ossification, new technologies keep getting deployed on the Internet.

For example, we have had IPv6 come in.  At some point we ran into a situation where there is more than 4 billion people on the Internet, and then we had to kind of get around it.  And it takes time but we are able to build newer things on top.  And these have technologies on the Internet now that the Internet pioneers couldn't imagine. 

The way we can put newer things on the Internet and expect them to work with people around the world is because of the openness and connectivity that's there.  It's something we should strive to preserve, like you said.

And so the other kind of key thing in there is the layering principles of the Internet as well.  So the Internet kind of holds together layer three and four of the model in a very high level and there is applications that we have a rich variety of applications but as long as we keep the kind of technologies and the lower layers so I would say a globally interoperable minimum, I think things are going to be good.

And that's what we should also look for, and also try not to push in, I would say, so I think like it was talked a little bit about the content in there, right, so the question is should content filtering happen at lower or higher layers?  I would posit it should happen at higher layers.  We are talking about transport things staying connected while enforcing millions of laws, state laws, country laws, and local laws are very different around the world.

So instead of trying to do this at a lower layer which the whole world shares, we should try to keep it at the high layers where it belongs.  That's also alluded to in the document.  One of the things like the messaging was giving us an example a lot.  We have something positive happening recently in this space with the multi‑stakeholder architecture, is that the Europe cape up with this digital markets act which got the gate keepers to open up communications and we started work on something called MIMI which is interoperating at the message layer.  This is a good blueprint to follow where the Governments and the policy organisations and the technical communities, we all work together to have these common goals of increasing the openness of the Internet, and people being able to connect.

And for the measurement, I think that's a critical piece, and I think we need to put a lot more effort into it.  And we need to have a lot more measurement points across the globe and be able to have a platform where people can use.  It's not just for us to do stuff, but also build platform such as a platform that exists today where people with run their own experiments on the platform.  So maybe we should let other people with IDS to measure things but use the same platform to build their own metrics on how they efragmentation.  Instead of us prescribing metrics.  So that is something that is actually really good as well.

I'm totally with you on the multi‑stakeholder approach.  I think it has worked well to bring the Internet to this level, and I think we should really continue going down that approach to work collaboratively and make sure that we learn from the past lessons we've had.  That brings me to my last 20 seconds to critique it, and the critique is really we kind of need a little bit more references out of this document.

So talk about Odyssey 1958 which talks about the future of the Internet and principles.  I think it's interesting reading for people coming from the policy sphere looking at what are the technical things that led to the Internet being the way it is and why it's very good for growth.  I think that's probably going to be my only critique on this.

>> SHEETAL KUMAR: Okay.  Thank you so much, and thanks for joining us online.  That was really useful to get your feedback on that chapter, but also you made connections to the other chapters as including the user experience and that's also key.  We do see different elements of the framework as intersecting.  The point is to help to provide a lens by which to have this discussion.

So if you all have comments on that, please do, of course, share, and you also, I think you made a point about referencing, about clarity of terms, about definitions which we also got in written feedback.  So that is something we can certainly incorporate.

But I will turn over now to Bruna, and Bruna will be facilitating this part of the discussion which is really to hear from you.  And so please do get engaged, and we will also be looking at the online participants for questions and reflections there.

>> BRUNA MARLINS DOS SANTOS: Thanks so much.  Yes, as we said, this is the feedback moment of the session.  So any questions or comments you might have are very much welcome.  We have some microphones in the room, so if you want to add some thoughts or just ask questions for the panelists, you can come to them.  I guess I will start with one remote question who is from Folye Habit from Togo.

His question is I would like to know how we can reach, how can we reach every citizen in the world, especially how can we overcome language barriers if content can be translated into our local languages, that would be very good.

And also he makes the comment about the more people are aware of splinternet damages and danger, the more they will be ready to fight against splinternet and defend their right to get access to available Internet.

My suggestion would be to take three rounds of questions, like three questions in one round and then I'll divert back to the panelists so we can start there.

>> AUDIENCE: I'm also a member of the Internet architecture board.  It's more a comment.  I would like to comment on the technical fragmentation a part.  Olaf talked about interconnect.  You mentioned this 50% definition from Hideman, which is really the 50% makes mathematical sense because if you have less than 50% it means you have the Internet and you have another network which is not the Internet which is just not connected, but at 50% you have two Internets, you don't know which one is your Internet anymore and there is nothing like two Internets, there is only one Internet.

This is very mathematical.  That's the point where it breaks of the that is the point where there is no way to get back to one Internet.  And we want to get there, and I think we are not on the way to get there, but if we get there, it's really done, then it's broken.  Then it's too late.

But what I wanted to say is that it's not only about interconnectivity.  It's also about the ability to innovate and evolve the Internet.  So if we put barriers into place where we cannot introduce new protocols because all of the protocols the way the Internet is designed it has to be extensible.  We were able to introduce even a new product on the IP layer, and it did not break the Internet.  We are still interconnected.

This is like all Internet protocols are designed this way.  You have to have a way to evolve.  And if you put barriers in the way where we cannot evolve anymore, that's where we need, whatever fragmentation, like a very negative outcome because that means not only we cannot change technology anymore, we cannot adapt, make it more secure, more flexible, more faster or whatever, better performance.  It also means that whatever we do on top of the Internet will be limited because we cannot adapt to it anymore.  And then we are stuck.

All of the benefits we get from the Internet where we see this positive impact on society and our economy, and so on doesn't happen anymore.  That's the point where we are still connected, but the Internet wouldn't be as useful as it is today.

>> BRUNA MARLINS DOS SANTOS: Thank you.  Right in the middle can we get a second question or comment.

>> AUDIENCE: First, I want to thank the policy network that they put the Internet fragmentation into perspective and we understood today what is meant by Internet fragmentation.  Obviously we have three dimensions, the policies and procedures, the user experience and we have the technical part.

And that lets us understand the subject much better.  The policy and procedure level, first what gives us comfort is there is a general consensus and agreement that we don't want the Internet to be fragmented.  So all of our effort is toward not fragmenting the Internet and this gives us comfort in this matter.

Our concern here is that regions and sometimes on national level, there were policies of let's say three Ts or commitments that represents the interest of the region or represents the national interest and in terms of social and economic and I wouldn't say all of them, but whatever presents the interest of the socioeconomics of this region.

And it is there.  And so these commitments or these frameworks or these agreements or treaties represent the interest of these regions or these people, and there should be a time between saying this represented fragmentation represents the interest or the benefit.  Maybe this is something that we need or something that needs to be addressed.

At least in terms of if there are any regional or national arrangements, there is a certain level that they should not conflict with the overall of the unity or the unification of the Internet.  Going back to the user experience and Vittorio as an advocate of user experience, give us kind of a trust that the indicators or the elements that have been identified, the five elements represents really truly the user experience at least principles that we don't need to be harmed.

And actually when it comes to user experience, there is nothing regional or national.  Internet users should be all equal.  So in that terms, we need to have a global understanding of that this is the minimum of what is known or what should be a user experience.

Going to the technical side, thank you that you limited this to the interoperability, and thank you that you clarified the decentralization or lack of connectivity or choice is not considered a fragmentation.  What gives us assurances that the industry or even the technical community built all of its work toward interoperability, and this is something that at least we feel trusted it will continue, but, again, bringing the matter to digital divide, it means returning this again to the user experience, which is now a wide open issue, and this may have implications.

Why it has implications?  In some parts of the world this may, a controlled user experience means a negative aspect on the social status of that user and the social status of that user can reflect limiting his freedom.  And sometimes limiting his social freedom or limiting his economic freedom.  This, so from all of that really, why we have some arrangements on policies and procedures and we have some arrangements on the technical side, we are wide open on the user experience and maybe this makes the start of the dimension of user experience more important than going first to the policies or the technical side.  Thank you.

>> BRUNA MARLINS DOS SANTOS: Thank you very much.  Right here.

>> AUDIENCE: Hi, I'm Barry Liba.  Olaf referred to a comment I made the other day about harmful fragmentation versus the fragmentation that's part of the work.  What I said is harmful fragmentation is the part where it puts us in a situation that is not the way the Internet was intended to work.  I came up to comment about the sort of gray area.  We think of harmful fragmentation as something where let's say a service provider blocks its competitors, access to its competitors or a country blocks certain websites at the IP level or the DNS resolution level, that kind of thing.

I have had a conversation with a colleague who works for Meta about who, for example, Facebook causes fragmentation, blocking off the content that Facebook users can see from people outside Facebook who can't get to that content, and, of course, my Meta colleague thinks that's not fragmentation, that's just the way an application layered on top of the Internet works.

And one thing that he says is Facebook is not the Internet.  The Worldwide Web is not the Internet.  These are application layers that are put on top of the Internet, but from the user's point of view, that often is the Internet.  So this kind of gets to Vittorio's area of fragmenting the user experience.

So I just sort of want us all to think about that a bit more, these gray areas that change the user experience in other ways that we don't normally think of as fragmentation and maybe we should start wondering whether it is and whether it's good or bad, just more to think about, I think.

>> BRUNA MARLINS DOS SANTOS: Thank you very much.  Next up.

>> AUDIENCE: Hi there.  Thank you, Christopher Tae from connect free corporation and Internet 3.  We think the future of the Internet is having everyone own their own IP address.  I think up until now, there has been huge amounts of cost involved in creating infrastructure that has led to ISPs and others owning blocks of IP addresses and having the difficulty of getting these IP addresses.  By everyone being able to generate their own address through cryptographic keepers we can give everyone an IP address. 

We think there is something cool going on in Japan, because of the Government has implemented a law against NTT in the 1990's, so that they had the NTT owned the network, but they weren't able to become an ISP.  They have fundamentally created a country wide layer two switching network where all ISPs can enter onto the network.

What that has allowed us to become is become an ISP of individuals.  Every computer on the NTT network using our software can have an IP address and connect and build presence on the network.  I think there is an interesting thing about decoupling IP addresses from networks.  It's very hard to have individuals create networks, but you we think there should be a decoupling between the infrastructure, the actual hardware physically layer and the layer 3IP layer.

So we have proven that this is a possible thing.  I think that there is a lot of discussions to be had, and we hope to join in these discussions so thank you for your time.

>> BRUNA MARLINS DOS SANTOS: Thank you for your comment.  I didn't see a fourth line there.  I'm very sorry.

>> AUDIENCE: Hello.  First of all, I would like to thank you for the panel and the report and the work of the network in general.  My name is Laura Perreira, delegate from the Brazilian youth fellowship, and we know that the defense of democracy, integrity is one of the main fields to actually adopt a more protective view of the digital space currently and in that sense sometimes to cause fragmentation and to risk the integrity of the digital space in general.

Actually in the Brazilian chapter of Internet Society we made an experimental application of the proposed concept of user experience fragmentation to collaborate on public consultation on Brazilian staring Committee on a regulation alerting to the unaverted risk of platform regulation when it does not consider this kind of harms to the critical properties of the network.

However, as mentioned by your presentation, it's not easy to balance democracy and integrity defense overall in the general sense and harmful fragmentations?  Is it possible to reach this sort of balance by using the concept of user experience fragmentation?  Do you intend to advance on this perspective?  Is it a goal of the network?  How do you see this issue in a more detailed way?  Thank you for your presentation.

>> BRUNA MARLINS DOS SANTOS: Thank you.  Just reflect that I'm closing the queue but we are going to take the last three comments.

>> AUDIENCE: Thank you for the panel, and I appreciate the discourse on Internet fragmentation but also just the difficulties surrounding understanding it.  And so I can keep it very pointed to the discussion points that were listed.  I am curious as we progress with initiatives like this, do we continue to do so without engaging regional or cultural leaders in areas that experience shutdowns or at the very least massive hindrance to their Freedom of Information? 

There was a point where national Governments is what we are hoping to interact with and no new stakeholders to involve in governance, however, there does seem to be valuable parallels between looking at the way that community have been oppressed in the past have taken a stand and helped create legislation and national policy to curtail that from happening to any other group.

I would like to raise the discussion group of meaningful connectivity as alluded to by the UN development goals.  With the rise of satellite availability and private corporate satellite availability Internet and LLM sophistication, do we recognize the potential for not only fragmentation, but quality of online experience?  Is this something where we see fragmentation leading from billions of people being priced out of meaningful connectivity?

And does this appear to be a perfect storm for exasperating the digital divide, not necessarily closing it.  So how does Internet fragmentation, policy design, how does it design itself to effectively account for rapid developments on these emerging fronts taking into account their incredible potential to create disparity of access to meaningful connectivity.


>> AUDIENCE: Thank you.

My name is Michelle Lambert, I work for an organisation called Equality, which is doing technology to support freedom online.  This is my first participation to the policy network and I'm particularly interested by it.  Hopefully, we will manage to create some governance that will prevent fragmentation.  I come from a background where we tend to believe that these discussions are difficult and sometimes they take more time, and we need to develop alternative technologies.

So I would like to use this floor now just to invite people to join us in Montreal.  We are organising a Conference called the splinter Com and the idea is to bring together the people developing new technologies which will allow us eventually to build bridges or make holes into walls so that people can continue to enjoy the Internet of so if you are interested to be part of that process, please go to SplinterCom.net and join us in Montreal in December to develop those technologies.  Thank you.

>> BRUNA MARLINS DOS SANTOS: Think you very much.

>> AUDIENCE: First of all, thank you for the excellent discussion on Internet fragmentation.  I have a general question, we want the Internet to be inclusive and open, borderless global network which gives equal opportunities to everybody who is connected and those to be connected, but because of geopolitical situation, trade concerns and other factors, the responsive nation states, some of the nation states take certain actions and enact certain laws that make internet digital sovereignty and which may result in technical, commercial or governmental fragmentation of the Internet.  So my question from you would be that how do you, I mean, you've given very good recommendations, but given the governance structure of the Internet how do you see how easy or difficult it would be to address this challenge of implementing those recommendations and especially we see the Internet evolving.  It's becoming more decentralized, so how do you see addressing that particular challenge and we talked about the five principles, the DFI principles.

So if there are certain laws enacted which may compromise those principles.  How do you see addressing that challenge?  Thank you.


>> AUDIENCE: My anymore is Raul from Latin America Internet association.  I think sometimes we are in a loop trying to define what this Internet fragmentation or not, and it makes me remember when we discussed it in the past about network neutrality and somebody introduced the concept, but we never had an agreed definition of that.  So we use a lot of energy discussing about this network neutrality instead of discussing what we want to do.  And if we look at the topic of this, of this event The Internet We Want. 

So trying to define what this Internet fragmentation, we have to focus on what things we want not to happen.  So I still, I think that the work the policy network is doing is impressive.  It's very good, and congratulations for that.

I have been part sometimes the discussions, but we should focus also more clear recommendations.  For example, for Governments, don't block apps, don't adopt policies that create different experiences in Internet for users in the same country, or in the globe.  This is the kind of things that we have to recommend, that, of course, I heard what the colleague said about not, about what we don't participate in a platform or in a given space, we don't have access to the information that is there.

But the point is that if I want to be part of that, I can, but in some places or due to some policies even if I want to be a TikTok user or buy something on Amazon or whatever, I can't do it.

So this is fragmentation, but so we will be in a loop if we try to say, okay, this is fragmentation, this is not.  But there are things that we clearly don't want to happen because that's the Internet we don't want.  Thank you.

>> BRUNA MARLINS DOS SANTOS: Thank you.  And I'm going to need to apologize for the two of you because we have closed the line and we have a deadline to leave this room at 10:30 at the same time we also have the process for bringing input to the discussion open until the 20th of October, but we really want also the panelists should be able to comment on that.  So, Olaf, Jordan, would you like to add any thoughts to the questions and comments?

>> OLAF KOLKMAN: Yes, well, not a lot.  I think that a number of points were made that are relevant and critical.  One of the points that was made was by Barry Liba asking for more nuance, and I think that's a fair comment.  We are also trying to keep this policy papers brief, and when you do brevity, you sort of lose nuance.

Mira made a good point, the ability to innovate and evolve is one we should protect.  That is indeed the idea.  I made reference to the critical properties, and one of the critical properties that we have defined and that we sort of introduced also in this paper is having an open architecture of interoperable building blocks, the idea that the Internet consists out of building blocks and protecting that open architecture whereby we can evolve.

I think that's important.

That sort of thinks to the gentleman that I forgot his name but I know from the metro advertisement, he invented something new.  I don't know if that works.  I don't know how that will scale across the Internet and as Mariel also pointed out we did the transition from V4 to V6.  That could have failed.  There is technical fragmentation between V4 and V6.

And the onus has been on the people who developed and are implementing V6, and give everybody their own IP address because that was the intention of the V6 address space to make sure that that interoperability from the V4 Internet continued to exist.

And that has been 20 years of hard engineering work.  Introducing something new will mean that the onus is on the entities that are introducing something new to make sure that that interoperability exists.  The critical properties says there are common protocols.

They don't say it's IPv4, IPv6, yet another protocol.  The Internet should be able to continue to evolve.  But we have to agree on something to keep that interoperability going.  Finally, the comment on meaningful connectivity.  When I talked about that evolution of the edge, this is a point we are making on the paper under the name of death of transit, the idea is indeed about having meaningful connectivity.

If the Internet evolves in haves and have nots, then there will be fragmentation too.  And being priced out of the market is indeed a way of being fragmented, and mind you, we have a fragmented user experience nowadays.  There are many people who cannot afford being on the Internet.

I think that's something we all have to work on, making sure that people who want to connect can connect.

>> BRUNA MARLINS DOS SANTOS: Thank you very much.  Ros.

>> ROSALIND KENNY BIRCH: Thanks so much.  That's actually a great transition line, Olaf because I wanted to come in on some of the first comments about local languages, for example.  I think this goes back to the broader thematic point we tried to capture in our chapter talking about the importance of inclusion in global Internet Governance bodies.  And I think local languages, making sure people can participate despite their cultural or regional background is so important.  So I really wanted to pick up on that point in particular.

And there were further points about particularly regional context and what not.  Absolutely.  I think, I really wanted to highlight the role of the IGF's national and regional initiatives in this regard.  I think these are great multi‑stakeholder spaces where people can come and talk about those local nuances, regional contexts, absolutely.

And I think better coordination between Internet Governance bodies as we have been talking about can hopefully help capture and bring those different voices together as well.  So not only just having these regional spaces, the NRIs.  I was lucky enough to attend the Africa IGF two weeks ago which is a fantastic opportunity to hear about some of these perspectives, but also to make sure that these are captured in the broader global discussions within these global Internet Governance bodies as well.

So thank you so much.  And just to say in general a big thanks to the audience for the participation here, and please do if you think of anything else, feel free to grab me on the sidelines throughout the rest of this week.  Thank you.

>> VITTORIO BERTOLA: Please join the discussion.  First of all, I think some of the comments pointed out what is, the problem we had to deal with when discussing the user experience level which is the user experience level is mentioned fragmentation is bigger event, as big as the planet and people only see a tiny bit of it and believe that that is fragmentation.

So if you talk to people from Silicon Valley, from the U.S. west coast they complain about what Governments are doing in other countries or even in the EU with the privacy laws, whatever, and if you talk to my friends in Europe, they complain what the Silicon Valley platforms do.  So the first time is agreeing on whether something is a problem and why and starting to work on that in a pragmatic way because if we focus on definitions, we will not go anywhere.  The other thing I wanted to say is in the end what we are facing now is the tension between the original dream of a united planet, borderless, everybody talking freely and the reality of differences of values, interests, economies and whatever language throughout the planet.

So to a certain extent you do need to preserve the local level and national sovereignty because that's a way to preserve the independence of people.  And to give them away, to give each citizen of the world away to have influence over the network and not just give it to the people that manage the network globally and have more influence on it, but on the other hand, you have to avoid breaking the globalness of the Internet.  So this is what we have to be concerned about, finding a balance.  Thank you.

>> SHEETAL KUMAR: Thank you.  I'm sorry we don't have time to provide the commenters from the earlier part of the session to respond, but the good news is there is still time to respond after the session by E‑mail or you can talk to us and we are giving a deadline of the 20th of October, and you, of course, have time to look in detail at the paper online, and the slides will also be available.  I think they are nicely summarize the in‑depth work that has been done.  So what we wanted to do, what the original mandate and detention of this policy network was to provide clarity to an incredibly complex and indeed controversial topic.  I hope that you agree that we have to some extent done that, but it is not over.

It is evolving just as the Internet is, an evolving framework and evolving piece of work.  Please do join us in continuing that work, and I think that is it apart from thanking you all for being here, for your contributions to the panelists, to the drafters, to the very active members of the network who gave their time to put this paper together.  Thank you.

And please do continue to be engaged.  We will be here during the IGF, but you can also email us.

>> WIM DEGEZELLE: Maybe if we can just have the slide because there was the link to the page.  Okay, and there we can see because on the Web page of the PNIF there is a link to the discussion paper and there is also explained how you can react.  So looking forward to your comments.  And the only thing I want to add is thank you to everyone.