IGF 2019 – Day 1 – Estrel Saal B – DC - IoT And Zero Rating Challenges For Net Neutrality

The following are the outputs of the real-time captioning taken during the Fourteenth Annual Meeting of the Internet Governance Forum (IGF) in Berlin, Germany, from 25 to 29 November 2019. 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 to understanding the proceedings at the event, but should not be treated as an authoritative record. 

***

 

>> MODERATOR:  Good morning, everyone.  And welcome to this annual meeting of the Dynamic Coalition on Net Neutrality.  This year will be on 5G, IoT, and zero‑rating challenges for Net Neutrality.  My name is Luca Belli, I am Professor of Internet Governance at FGV Law School, and I have the honour to Chair this coalition.  This year's program has been dedicated to these two relatively new topics that are emerging as Net Neutrality challenges on one hand, price discriminating, zero‑rating applications so sponsored applications that are not counted against the data caps of users.

And on the other hand, the emerging 5G technologies which are also creating some Net Neutrality challenges so I go like to introduce the panel, we have Aurore Tual, the Special Rapporteur for the Organisation for American States,.  Frode Sorensen from Nkom, KS Park from the University of Korea, in South Korea, not North Korea, and we have Thomas Lohninger from Epicenter Work and then last but, of course, not least ‑‑ we have Bob Frankton, but he is not speaking on behalf of any IEEE things, and last but not least Veronica rare rose from Access Now.  So we have a why crowded panel for these two very relevant issues.  We have decided to structure this into two different segments right after the introduction about Edison.  The first explaining by zero‑rating so sponsored applications are having some positive sometimes but also a lot of negative externalities on how people enjoy Internet access especially in Developing Countries where people do not have the money to pay Internet access fees and, therefore, they primarily access sponsored applications and this has, of course, an impact not only on the fact that these applications become the main vector for information or disinformation of people that, for instance, receive fake news primarily through sponsored social networks.

On the other hand, it has also a another important negative externality which is that not only content is vacated through sponsored applications but pearnl data are extracted primarily by those dominant usually applications.  So this has, of course, effects not only on the economy.  It's more difficult for new entrants for start gyps to enter the market if only the big applications are sponsored and the others have to be paid, but also in democracies.

If data, personal data of people are collected and centralize the only by a few dominant applications then those people can be profiled and targeted content can be sent specifically to them on their mobile which is, of course, the main Internet access venue nowadays not only in Developing Countries, but also in Developed Countries.

So these are only part of the concerns we are going to discuss today before we enter into the 5G debate.  So without further ado, I would like to invite Edison Lanza to introduce the debate of today with his keynote remarks.  Please.

>> EDISON LANZA:  Good morning, everyone.  Thank you, Luca, and the Dynamic Coalition of Net Neutrality.  It's a pleasure to be here and to make some introductory remarks.  I'm the Special Rapporteur of freedom of expression in the Inter‑American association of human rights that has the mandate around the atmosphere, but we have a common, you know, approach with the special Rapporteur of United Nations and I think that under the international law, the principle of the Net Neutrality is Internet design principle, my office remarks on different reports and different statements that Net Neutrality is a necessary condition for the exercise of freedom of expression and intersect with the guiding principles.

The purpose of this principle is to ensure that free access and user close to use, send, receive or interfere any local content, information, ideas, and to choose applications service through Internet is not subject to condition or directed or restricted, such as blocking, filtering or any interference.

In the, in our region in Latin America, several countries have already enacted enforcement and established the principle of Net Neutrality in their framework, legal framework, Argentina, Chile, Mexico, Paraguay, Ecuador have some disposition that ensure and oblige the service provider to respect freedom of expression and, you know, the Net Neutrality in the management of the data packet.  And also in 2015, we have engaged with the FCC policy on net neutrality that at that moment was the most powerful in the region that protect and ban the practice of blocking or restricting what the people can do or see online, prevent threatening, a specific provision that the guarding of traffic by source and nation and content is preclude and ban the pay prioritization.

In 2017 the current Government in U.S. and the current commission of the federal communication of communication take down this rule in U.S., but now they are fighting against this decision as many state in U.S. now pass law to enforce the principle like California and New York and others, and it's now in a problematic situation, but I think it's important to follow the decisions also in the U.S.

In respect to the zero‑rating policies, we release our special report about that, and acknowledge that the new and controversial debate arose in 2015 regarding these plans that zero‑rating plan that allow Internet providers to provide access to a specific application without that access being charges expended in the end user data plan.  And if I say that many countries have and protect the Net Neutrality rule, also I want to say that in many cases in Latin America, there are an exemption to commercial plans, and in fact, you know, all of these countries, Mexico, Ecuador, Brazil, Argentina, has allowed the practice of zero‑rating in all of these countries, and in fact there is some research about the civil society organisation that alert the people in Latin America do not have their knowledge if they don't have access to the whole Internet in some of the cases.

The approach of the zero‑rating plan is different.  In some cases it's a free access all of the time to some of these applications.  In other cases it's a packet of data and when the packet of data finish, the people want to and can access to a number of application.  Finally in regard of the impact of externalities of these plans in Latin America, I think that in one hand this kind of application could improve access of people that have, you know, the level of money to access this, but in the other hand, we have a huge problem with the universal access in similar conditions for other people in the region, and in the international law, the states have the duty and the obligation to ensure the whole access to Internet for other people, and this kind of planned on, you know, substitute this obligation by the state.

Finally, I want to say that in the context of this politicization and political, politicization environment in Latin America, we acknowledge that in some cases like Brazil, this kind of plan has a very problematic impact in democracy and in the access of information in the context, in the intellectual context.  Why?  Because in many cases WhatsApp, and this kind of application ‑‑ Facebook and this kind of application is only one application that millions of people access in the context of, you know, and whether they use this information, this information flow through these platforms, this population don't have the possibility to check out with other information or other services the information that flows through these platforms.

And for cases of climate change and other issues this is a huge problem that we start to look in Latin America, and it's very useful to know what things are in other regions of the globe.

>> MODERATOR:  Thank you very much, Edison, also for highlighting that this actually in the Americas, especially in Latin America, these plans are existing are allowed in spite of the existence of Net Neutrality policies.  So it is a choice, explicit choice of the regulators in those countries to allow this kind of plans based on the calculation maybe that is better to have a little bit of Internet than not having it, but probably this is also based on a miss calculation that when you have a little bit of Internet, that little bit becomes much more easier to manipulate and be used to extract very useful information and personal data from those countries, and this may have a nefarious also impact on democracies besides the economic impact on competition.

Before we start the first segment, I would also like to clarify that unfortunately we do not have any representative from the telecom sector today.  I invited four, but none of them unfortunately could participate.  But I'm sure there will be some representative from the telecom sector in the room, so whenever they want, they are absolutely invited to provide their feedback and input and help us having a more constructive discussion.

So now opening the first segment on zero‑rating, I would like to ask Thomas to provide some insight based on the study they have been developing.  He has also presentation.  While the presentation of Thomas is put on screen, I would also like to invite you to check our annual outcome of the coalition, which is the website zero‑rating.info to provide you information on zero‑rating as you may guess from the URL and it consolidates basically a lot of resources on zero‑rating plus a zero‑rating map that provided you information on which kind of applications are sponsored around the world, and which kind of policy Net Neutrality zero‑rating policy exists around the world.

So this is a free to use resource that you may access and share as you want, and hopefully also regulators will use to have more informed policy, which is our final goal.  Is the presentation of Thomas ready?  Excellent.  Thomas, please go ahead.

>> THOMAS LOHNINGER:  Okay.  Wonderful, okay.  What I'm going to do now is to briefly explain where we are right now but first I should introduce I'm speaking been highway of the of Epicenter Works it's a Vienna bases digital rights organisation and I'm Chair of the Net Neutrality Working Group in European organisation of other digital rights organisations in the European Union.

Most of my examples are from the EU, but I think some of the conclusions could apply globally.  Last year in Paris, I showed you this slide which were the preliminary results of a study that we did in which we basically analyzed and looked for all of the zero‑rating rate offers in the European Union and economic area.  So basically all of the countries where the existing Net Neutrality framework applies to.

You can see that five of the top ten applications are from one company, from Facebook, and on the free applications in the top 20 are based out of the European economic area.  And again another analysis that we did, you request see here, we basically looked at the zero‑rating offering in these countries, we could identify 186 offers.

The data as well as report is available under free license on our website, epicenter.works and here we analyzed degraphiccal relationship between the ISB and the head quarter of the provider of that application service.  You could see that in the normal type of zero‑rating application, zero‑rating program where the ISB is choosing the politics that participate, we have a strong tendency towards U.S. based dominant applications profiting from zero‑rating.

But there is also a new type of zero‑rating offer being drawn from T‑Mobile in the U.S.  It was the first and many more of these open programmes can also now be found in the European Union.  They distinguish themselves by allowing some form of sceneup for interested information providers to join the zero‑rating program, and if you only look at these, you come to the picture that there is actually a strong benefit for local applications and then in second place you still have applications with the head quarter in the USA, and then particularly from a European perspective, the cross‑border provisioning of applications actually drastically goes down.

And we then further dived into that data set and analyze the how many of these zero‑rating deals can an application sustain and we found that the majority of applications only can sustain between one and three of these deals with telecom company and that, of course, is because there is first an administrative burden, you need to know about the program, you need to sign up for it, you need to enter into commercial agreements that often come with a liability for wrongfully built data volume and you have to supply identification criteria to the telecom company to make your service distinguishable in the network of the ISB from other traffic so it can be belt differently.

And basically, every application has to enter into a contract with every ISB whose customer is might want to reach or serve a competitive offer.  And and the peak there at the end between 31 and 52 customer signal agreements, that is the top 20.  So the whole study is available online on in URL.  It was released in January and it is our contribution to the ongoing Net Neutrality reform in the European Union.

I think we also want to talk about 5G so I brought a little bit of basic examples on the new mobile network generation standard.  It is really important to stress that 5G is not a revolution but an evolution.

So like previous mobile network standards, it is not drastically changing everything.  One particular technology aspect that we have to be cautious about is network slicing, and as the name suggests you are slicing up the network into low latency, high bandwidth or low energy consumption slice and the use cases that the industry is putting forward in order to justify why 5G is so drasticically different, one is self‑driving cars.  It's often said that they rely on 5G network in order to function.  We dispute that because then these self‑driving cars could not go to rural buttenburg outside of Berlin or to the mountain area and that's why the European Commission backed a WiFi based standard for car to car communication.  It's a mesh network, and if you then look at the statements of the consortium that vector 5G standard it is quite funny that they are complaining about discriminatory interoperability and compatibility requirements.

Another example is you hear about is connected farms.  There again I would dispute the basic premise because rural areas right now have the worst mobile coverage in most countries and also the least amount of people living there, so it would probably be economically unwise to really invest in these areas although it would be beneficial for the people to live there and invest in the most expensive modern technology should be questioned.

Another example is industry 4.0, connecting facilities and here I will bring an example of Europe.  The German association of chemical industry actually applauded the German regulator about their decision to hand out part of the 5G spectrum for independent use.  So if you operate a factory in Germany, you can just apply with the regulators and get that spectrum for your specific location and build your own 5G network.  The chemical industry in Germany is very happy about that move and it justifies this with the confidentiality and integrity of their data.

Lastly, remote surgery over 5G that's a real example from China.  And to even bring this up is kind of missing the point because you would never do remote surgery over the open Internet.  The open Internet is not capable of delivering the levels of assured quality so you would go for a specialized service for that and keep it as far away from the open Internet as possible.

Finally, there is a case to make in some extent in online virtual games because there is a low latency slice.  This is also a discussion that should be really technical, and I just want to debunk this a little bit with the numbers of the telecom industry.  This is GMSA the global association of telecom operators, and even their most optimistic projection sees a 40% share of 5G mobile connections by 2025, so with a market share of such gaming consoles is quite small.

And lastly, an example of Internet of Things, IoT, and, yes, here I would also really caution the type of economic incentive and ecosystem that we create because if a certain type of device really relies on a specific type of electric slice connectivity, you end up with this.  Instead of a universal connection that is the neutral Internet right now or a universal power plug, country specific, of course, you end up with you very application specific technologies and I don't think this is from a sustainable perspective the right choice to make.

And because I'm over time, I will believe it at that.  Thank you.

>> MODERATOR:  Thank you very much, Thomas.  You have provided very useful inputs for the next segment on 5G.  Now, I would like to ask to KS park to provide also some inputs on what are the most recent developments of the Net Neutrality debate in South Korea.  South Korea is actually one of the few I would would say two, maybe three countries that is massively investing in 5G, but it is a country where a lot of issues that we have been discussing over the past years about traffic discrimination or prioritization or any kind of commercial agreement between different vertical integrated actors is becoming much more visible in the, over the past couple of years.  So, KS, please go ahead and provide us insight on the South Korean debate.

>> KS PARK:  Thank you.  I think we should first establish that we love Internet because it makes, it makes this project, it makes this project scalable.  The project of letting everyone, even policy individuals to become the actors in mass communication where people can connect with one another without, without receiving approval of gate keepers like newspaper or television, having said that, Net Neutrality is crucial to maintaining that significance of the Internet.

I think it was a mistake for Tim Oh who coined the term Net Neutrality to use public utility or other common carriers like gas or electricity to promote his idea of Net Neutrality as a non‑discrimination principle because gas, electricity, these are all subject to usage‑based pricing.  Where people pay for what they use.  Now, in Internet, there is no provider of any services.  There is no receiver of services.

The idea behind the Internet is that, the idea behind the Internet is not to create a gate keeper, not even a reasonable gate keeper so that all of the computers can talk to one another without actually connecting with one another.  Now, setting it up as a non‑discrimination principle leaves open the possibility of allowing reasonable discrimination.  Now, there is no such thing as reasonable discrimination on the Internet.

The Internet is basically a promise, a promise that all of the computers will receive packets from one neighbor, pass onto another neighbor that is closer to the destination that is marked on the data packet free of charge, free of any non‑functional condition.  If that is the Internet, that means all of the computers are participating in the delivery of data for all other computers.

So I think that Net Neutrality should be re‑established as a principle that there be no financial or non‑financial condition for delivering, for delivery of data packets.  Now, it's not my invention.  If you look at the economic, if you look at economic literature, the way that economists define Net Neutrality is that termination fee is equal to zero, which means there should be no money demanded or any other condition that data packet constitutes certain content or type of content in delivery of data.

So why am I talking about this?  Well, if you look at, I mean, the idea that you pay for what you use, this is so hard to shake off people's thinking.  So in Korea, if you look at Korean Net Neutrality guideline, it allows reasonable discrimination of data packets, which means that whenever, whenever mobile Voice Over IP packets are throttled or blocked by local telecos who want to protect their voice revenue, you know, normal telephone calls, whoever is making an argument against that has to do all of this economic analysis to somehow win over to prevail over the teleco's argument that they need to protect their voice revenue to be able to invest in network building.

So plaintiffs have the burden of proof who they could not use the access they purchased for the use that they want to make of.  And another aspect of network neutrality guideline is that it more freely allows network slicing that Thomas talked about because the way it is phrased, as long as the minimum quality of general Internet is protected, the telecos can use the surplus bandwidth for whatever purpose they can make use of, for self‑driving cars, they can slice up the bandwidth and sell each slice at a higher price as long as the minimum quality of Internet is protected for for ordinary consumers.

And this loophole in the guideline, again, comes from this idea that there is, there has to be payment for delivery of a package.  So I wanted to talk about Korea but also pick your brain about how to re‑establish Net Neutrality from the fundamentals of the Internet so that it becomes a better functioning norm on the discussion on 5G network slicing and when I have more time, I will talk about usage‑based pricing that has become another really restrictive norm in Korea.  Again, when I have time, I will go into that.

>> MODERATOR:  Thank you very much, KS for respecting your timing.  Indeed you have raised relevant issues which is the fact that in many countries although there has already been elaboration of Net Neutrality guidelines policy, we actually say that there is a divergence between what is in the policy and implementation, in the practice, and also we see that many of the issues that we took for granted, the fundamentals of Net Neutrality like no discrimination based on commercial agreements are actually not really so respected or so shared even when they are included within national policies.

Now, to conclude this first segment, I would like to ask to Veronica from access to provide some elements also to have a better idea ever what also Edison was mentioning before about the wide share adoption of zero‑rating practices in Latin America, and how this is having a concrete impact on Latin American societies.  So, please.

>> VERONICA:  First of all, thank you very much for this invitation for us it's quite important to be mere in this Dynamic Coalition session on Net Neutrality.  Access Now has been fighting prying to promote Net Neutrality principle in every space.  Today I want to take some time to talk about two initiatives that are happening in my only country, I'm from Peru, and I wanted to talk about that.

We have Net Neutrality regulation there, however, you might remember free basics, right,, we talked about that for many years.  But nowadays it is trying to launch a free base 2.0 which is essentially going to be a browser, they sell it as a browser.  So the user is going to download this application, and through this application, they will access to free content, all of the Internet will be free.  They will not have access to videos nor images because you know that that costs, and already this is being implemented in Peru and in Colombia, and they are going to spread this in the next months, but they are trying to relaunch this, I don't know exactly when, but we have been having conversations with Facebook regarding this point.

First of all, we have to say that these new type be this new generation of Free Basics kind of listened to our claims when we were discussing first free basic at that time we criticize violation of Net Neutrality principle, the security of the initiative, also its position as global gate keeper for Internet connectivity., however, nowadays with this new episode of Free Basics 2.0, which is there and people can use it now in this moment.  They have not much users now, but I think they are expecting to increase the quantity of users, but we want to raise some flags here.

First of all, there is lack of information of how it works.  At least in Peru, it works with ISB one is intel and one is Btel, a Vietnamese company.  There is no information how it works, there is no information to the user, so it's available so the user can download the app and start using this.  There is no information regarding the quantity of free data they have every single day.  There is no way they can check how much data is left.

There is no information if the ISP or Facebook can check the traffic or how they are doing this question.  We hope and we expect Facebook is going to release more information soon when they relaunch this application.  Also there are problems with security because we already ask Facebook for the specifications to get a look on how this exactly is working because we don't know exactly how Facebook is managing the traffic use in this browser, again, I'm trying to use this because I don't know exactly how to explain it.  They sell it as a browser.

What happens if this user wants to use the browser for banking operations is the entire Free Basics 2.0 encrypted or not?  We don't know.  We don't know exactly if they can, as they are going to be like the door to the Internet, if they are going to discriminate or give some references to one service to other service.  We don't know exactly that.

Also I do believe that this creates a perversive incentive as when we were talking with Facebook about this, they show us that among the most popular services that users access through this browser, they have YouTube, Facebook, and InstaGram.  As you know, the experience of the users in these platforms is mainly videos and photos.  That's exactly what you are doing InstaGram, right?  And then if Free Basics 2.0 does not allow you to watch videos or photos, then how is the user loving this application using Free Basics 2.0?  We don't know how that is happening.

What we see is they are giving this free sample of Internet which is a really bad sample because it doesn't allow you to see videos and photos but then you increase desire to watch those things and they start to buy data because they have to use their data to watch those things.  This is good for the ISP because you get more uses and you can sell more data.  Another worry that I have here is that if they are selling this as like a browser idea, isn't this something to worry regarding vertical concentration maybe because they have this, they are in many layers of the Internet, but now they are trying to also be there in the browser market.

When I see this thing happening in my country, at the same time I see things like this happening.  This is like an ad of the Vietnamese national company which operates mainly remote CTs where you can access 2G, 3G Internet.  They have, this is a plan or a program.  They have zero‑rating.  Those are the applications, because even though we have a regulation for Net Neutrality, we do accept zero rating when there is no preferences among the service that compete around each other.  But this thing is that they give you with less than two dollars, they give you 1.5 gigs for Internet apart from the zero‑rating.

When you are done with your 1.5 gigs, you have accessed your free data.  And it's limited.  It says unlimited here, and you can read the specs here.  The only thing is that the speed gets lower, and but you have free access to everything.  So you have, this initiative, these kinds of initiatives happening and at the same time we have Free Basics 2.0 that is trying to sell this new concept of Internet without images and without videos.  So this is something to worry about.  I just want to put this on the table, because this come up soon.  We are still waiting for Facebook to be more open and I think they will do this when they relaunch this application, but this is something that is still there so people are accessing to this thing there in Peru and Colombia.  That's it.

>> MODERATOR:  Thank you, and before opening for the first round of comments and questions.  I think one important point has to be highlighted the fact that we usually consider zero‑rating only from the perspective of the recipient the customer that receives something free, a gift, free content, free video.  But the it great questions that one has to ask, and I hope Government and regulators will ask everyone unless we have been living in a cave for the past ten years know that data are the most valuable resource in the world, it is is this really free meaning everyone knows there is not such a thing as a free lunch.

So if you receive something free from a for‑profit corporation, it's quite hard to believe that this is completely free, and it is not paid with personal data.  And if personal data is the most strategic and valuable asset we have nowadays is if wise to give open access to national strategic resources to dominant and usually foreign corporations that have demonstrated over the past five, ten years that are not the most careful corporations to be trusted with people's data.

That is just a provocation that I am dropping there to start hoping that this will spark provocative comments and questions in this first round of comments and questions so if you have any, there are mics there that we can, you can use to add your comments or ask your questions.  So please, if anyone has any comments or questions about this first segment, go ahead.  The mics are there for you.  Don't be shy.

Yes.  I hope you could just take the mic yourself, because we don't have anyone helping with mics in this moment.  So if you can use that mic.  Thank you very much.

>> AUDIENCE:  Hello, my name is Abdul Shihar.  I'm a Ph.D. student in telecommunications law at the University of Strasbourg.  I just would like to get back to the Peruvian example.  It was very interesting what you were talking about, and I want to know exactly what are the concerns of this discrimination in connectivity?  Is it about distortions in competition between the companies or is it a social concern about the connectivity of end users?  Thank you.

>> MODERATOR:  Yes, we have other questions and it's better if we have two or three and we get back to panel.

>> AUDIENCE:  I'm from Ecuador.  In the same line Veronica you say about the zero‑rating applications.  In Latin America we have a unlimited but limited functions about A PP.  We have some functions about Whatsapp are free.  In the European case, it's the same.  We have some functions restricted or they are free or unlimited.

Thank you.

>> MODERATOR:  Okay.  I think we can get back to the panel unless there are other comments in this moment.  Okay.  Do you have a comment there.  Yes, please, if anyone wants to take this, if Veronica wants to start or Edison.  Edison, please, go ahead.

>> EDISON LANZA:  Thank you for the comment, and I think that a big problem in Latin America is that the regulators that have the capacity to monitor and measure this kind of plans do not have any, you know, concern about that.  And in the report that I mentioned, I said that the state needs to assist in the competitive of the policies with the term of the roles of the Government and the regulator of the Net Neutrality, and a test with the human rights impact in this case.

Because in some ‑‑ I think it's positive to try to improve access with very high inequality, but on the other hand, it's a duty of the state to have a measure and monitor the impact on human rights in regards to privacy, freedom of expression, social impact and democracy impact.  I think that, you know, I don't know any regulator that releases some report or some information about the impact of the monitoring of this this kind of plans.

>> MODERATOR:  Veronica.

>> VERONICA:  Regarding the question, okay, we are worried about, yes, we are worried about, we can always worry about competition.  There is concentration happening in telecom industries that is quite normal, again as we are not expecting to have that.  But also our worry is more related to our rights and that's why the first thing that I emphasized was regarding the lack of information that we have right now to exactly know how this new Free Basics 2.0 works because we are worried about the security of the application.  We are worried about how the data has been processed and how the traffic is happening and everything.

If this is going to be something that will be available for everyone because nowadays, this is just available in two ISPs, but they are working, and they are open to work with other ISPs as well.  So at some point that can be open for everyone and available to everyone.  So our concerns are that.  I don't know if I'm answering your question, but I can go deep on those points later after the session, yes.

>> MODERATOR:  There is also Thomas and KS wanted to add something.

I may just interject to also take advantage of my moderator role, to abuse of it and say also something about the fact that the question actually if the concerns are more about democracy or economy, it's a misleading one, because we should start understanding that the concerns on democracy and economy are intimately intertwined.  If you don't have competition you don't have startups in your country generating revenues, work, it means that people that are unemployed and also only use used to generate personal data that is exported abroad at some point will be furious and will vote any kind of clown that tells them that is going to restore order and prosperity, of course, line, and the two things are intimately intertwined.

The fact that lack of competition is a threat for democracy, because if you have less entrants in the market, if you have less business, you have less jobs, and you have more furious people.  So for social stability purpose, I would not suggest to any Government or regulator to reduce competition because people that do not have jobs and work still vote.  And so that means that at the next election they will have to meet them and they will be furious.

>> THOMAS LOHNINGER:  I think it's establishing how much we are all agreeing on this panel but I want to also add to the last point.  Of course, you could also look at the question on the cost of zero‑rating as the cost of innovation, because you actually increase the entry barriers to any country where you have zero‑rating and particularly if you try to extend your service beyond just one country or one region, if you really try to challenge incumbent players, you will find that these zero‑rating programmes no matter how open and non‑discriminatory they are will become an obstacle to grow.

And this can be seen by Netflix when it was obvious that the U.S. would roll back their regulation the CEO of Netflix said that shareholders shouldn't worry because Netflix is big enough to survive in Net Neutrality.  So in a way every company that doesn't have dominant platforms should be in favor of abolishing zero‑rating.  If the privacy cost, because in order to provision these types of services, what telecom companies often have to resort to is Deepaket inspection to look deep to the S and Is or URLs of every data package in order to count and build them differently.

Another question was asked by this gentleman about the fact that most of the zero‑rating programmes only apply to certain functionalities.  So, for example, WhatsApp is zero‑rating.  That doesn't include WhatsApp video call os are video messages.  And you can see arbitrary distinctions that are often not transparent to the consumer which functionalities, which applications are actually a part of the zero‑rating or not.

If you have research to support for yums of many of these programmes you will find customers being angry and not understanding their bill anymore and if you think about it, in fact the URL of every data package will become big data at one point because these big measurements are fight quine, granular, and they are bad for innovation because if you create a multifunctionalty app, for example, a messenger, it is a messenger but also a groupware and a file sharing software.  In which part of the zero‑rating bucket should it fit?

So it's actually a quite arbitrary distinction to try to chop up the Internet into individual pieces and then sell them separately, which is just not the way this was meant to be exercised.

>> MODERATOR:  KS.

>> KS PARK:  The main concern with selective zero‑rating is the destruction of competition between different contents.  Number one, teleco in Korea both in wired Internet and wireless understood is SK telecom.  SK telecom began online shopping mall, and zero‑rating, the only shopping mall, but no other shopping mall.  SK telecom never had experience in online shopping mall, but once they began doing that, that online shopping mall, you know how tough competition is between different shopping malls in any country.  SK tell coms affiliated shopping mall became number two and still has maintained that status.

There was a research on how many people are willing to change their teleco to use online shopping mall free of data caps.  65% said they are willing to change their teleco to use zero rated online shopping mall.

That shows how much lack of competition or how much disruption of competition can resort from zero‑rating.  Now, having said that, let me say something controversial about zero‑rating.  The pre‑Trump FCC of the United States under Thomas Willard, did an analysis of different zero‑ratings.  The paper although it was never officially published as a statement of FCC, said that zero‑rating, teleco zero‑rating their own contents is heavily presumed to be anticompetitive, but zero‑rating the independent contents is not.  So T‑Mobile zero‑rating was it Netflix?  I forget what the content was, but that was considers not anticompetitive.

Now, I think that there is some lesson there in line with what I said before.  The Internet is born free.  The Internet, you don't purchase data.  You purchase access.  So zero‑rating, I mean, the Internet is supposed to be zero rated to be born with.  Now, when things became mobile because you cannot predict how many people will be using certain access points we began ‑‑ let me just finish.  We began applying data caps, but data caps are not supposed to be there to begin with, and with the new 5 technology, maybe we can get rid of data caps on mobile as well.

So if zero‑rating allows, I mean, if some of the zero‑rating programmes allows to use Internet without data caps, the more the better.  Think about that.

>> MODERATOR:  Sorry, Bob, I see that you are boiling, but to stick to the program, very short sentence, please.

>> All of the zero‑rating is about symptoms.  The fundamental concept of telecommunication from 1870 or even 1830 is the problem, and arguably zero‑rating is arguing of minor symptoms on the side and we have to get to the real dysfunctionalty.  I will get to my talk.

>> MODERATOR:  KS was mentioning 35G it provides good segue to start the second segment which is going to tell us a little bit more about how our set is approaching zero‑rating, not only zero‑rating, but Net Neutrality violations and then a little bit more on how they are approaching 5G.  Thank you.  Can we have the presentation, and I would like to thank our remote moderator, and if there is any comment or question from the net, please tell us so that we can address it.

>> AURORE TUAL:   Thanks, from the invitation.  ARCEP is a French regulatory regulator.  It is architect and guardian of the country's network network, fixed and mobile tell coms and networks.  The authority works to ensure that those networks develop.  Regarding net neutrality as indicated and using different tools to understand genesis of compliance with the European open Internet regulation.  So firstly, monitoring terms and conditions, press review, but also social media.  Analyzing user reports through a platform called (?) discussion within the bar back with other rig fore authorities are also at full.  For instance, regarding zero‑rating, proactive dialogue when identifying person service restriction to Net Neutrality is also used.

For instance, ac set focused and freedom of choice and device in the ISP's mobile plans after ARCEP took action we limited the use of SIM cards.  ARCEP made available an application called Wehe it was developed by the university and can be accessed by any consumers who read an IOS application.  So the testing tool compares time it takes for traffic generating by certain services to be relayed.  It measures the difference between the traffic stream actual travel times through the network layers and the travel time for similar but encrypted traffic stream.

It's the reason for giving significantly different in a similar and matching fashion, then it is possible to suspect that the operator as has implemented measures that affect traffic.  User request then decide to inform ARCEP which will be in position to investigate these reports.  So this new distributed cool is part of ARCEP initiatives that are designed to help consumer making every every one a participant in the regulatory process.

So partnership between the university and ARCEP is still going on so as to adapt the services tested to the message one in France and to develop functionality for testing prioritizing of ports.  In France, more than 62,000 tests have been relayed so far.  As yet, none of the results provided by the application have made it possible to suspect any traffic management practices that violate Net Neutrality rules in France.

And finally, according to European and national legislation, ARCEP published every year a report on the state of the Internet in France.  And a summary is available.  I have a few here if you are interested.

So we are very happy to continue discussion on Net Neutrality and in particular on Net Neutrality and 5G.  So regarding 5G, ARCEP is in line with the position that conclude that open regulation is neutrality neutral.  It is not that the regulation seems to be leaving considerable room for the implementation of 5G technologies and it is not aware of any concrete example where implementation would be impeded by the regulation.

So based on this statement, ARCEP is involved with stakeholders to under the new use cases allowed by 5G technologies, and stakeholder needs.  To this end, France launched 5G pillar window for market players in two frequency bands.  So the first one is a 3.4 to 3.8 gig band, more than 70 experiments were carried out and most of them led by telecom industry players but also by academies.  And the other one is a 26 gigabytes band where 12 projects have been authorized including several projects led by verticals or consortia that do not specialize in telecommunication.

So the objectives of these are to identify the players along the swral you chain and in testing out business models, allocating frequencies to interested players to conduct the first 5G trial and obtaining initial feedbacks from the stakeholder.  Just two quick examples, concrete examples ftz ongoing experimentation in the 2.6 gigahertz band.  The one is it is national velodrome in France.  It's video recording of cycle cases with use cases that reaches from augmented reality of the different events to application by applying fixed and mobile in equipment and sports media.  This open tier platform could raise challenges facing future on MP game sites.  The second example is the port.  The goal is to explore and test 5G application in a port in an industry‑related context.  This includes application in the field of energy such as operating small grids or recharging electric vehicles.  Other applications will focus more on logistic applications in the port area like operating terminals.

In the ongoing experimentation, no question regarding Net Neutrality have been identified so far, but we remain attentive and open to discussion.  Thank you.

>> MODERATOR:  Thank you each, Aurore Tual and thank you for saying that regulators should partner with geek in order to have creative solutions to solve problems and we see that the results are quite impressive when this partnership happens.  So to have some other ideas while regulators are acting and how they are tackling 5G issues, the balance between 5G and Net Neutrality policy and what are really the 5G concrete examples and evolutions that we can witness.  I would like to ask Frode Sorensen to provide us a little information about what Nkom is doing and what are the concrete examples you are seeing in your work as a regulator.

>> FRODE SORENSEN:  Thank you very much, Luca.  In my intervention, I will discuss the compatibility between Net Neutrality and 5G.  And the intervention from ARCEP was a good start up to my presentation.  I'll go a little bit deeper into the 5G use cases and as a Net Neutrality regulation example, I will necessarily use the Norwegian and European approach to Net Neutrality, but I expect it to be applicable to Net Neutrality in general as well.

If you look at the 5G networks as they are marketed by operators and the organisations, they foresee three main use cases.  It's often drawn as a triangle where you have the enhanced mobile broadband in one of the corners, which means high speed Internet access simply, and then you have the low latency communications, real time communication in the second corner, and then you have the massive machine to machine communications in the third corner.

I will start to discuss Internet access services compared to these quality of service‑based services.  There is a major misunderstanding regarding how Net Neutrality regulation works.  It seems like there is an understanding that Net Neutrality means that you can't have good quality of service, and that's not the case.  In the European regulation, of course, we foresee a good speed on the access to the Internet itself.  But there is also a possibility to provide other services than Internet access services.

These services are often referred to as specialized services in European regulation.  Then the question is when can you provide a service as a specialized service instead of an Internet application?

So what the European regulation foresees is that if a service needs service and it cannot be provided over the Internet access, then it is still allowed to be provided separately as specialized service.  So, therefore, these low latency services that we foresee in the 5G networks is fully compatible with the European Net Neutrality regulation.  And more than that, it actually facilitates this use case because when you have the 5G networks fully developed, you will have this network slicing functionality which was touched by Thomas in his presentation, so this network slicing function actually helps to provide specialty services in a way that is compatible with the European regulation, because there is a requirement that the Internet access service is not degraded, but the general quality of the service should be kept up to speed when the specialized services are provided in parallel.

So, therefore, we believe that 5G will actually be easier to implement than 4G under the European Net Neutrality regulation.  Secondly, regarding Internet of Things which is in the third corner of this triangle, # # what is IoT?  What is machine to machine communication.  It is actually just an application.  It's an application running on aon a separate piece of hardware so the goal of achieving innovation of services, invasion of applications is innovation of applications is actualli applicable, of course, to IoT which is a kind of application which is running on its own hardware equipment.

How can that be compatible with European Net Neutrality regulation?  There you have a separate misunderstanding that is also disturbing, the discussion about compatibility.  Because Net Neutrality regulation does not regulate any communication in the world.  It doesn't regulate all of the communications we have, because there are also private communication networks, for example, a corporate network is not regulated by Net Neutrality.

The regulation is provided for public services.  So if you have a private service, it's not regulated, and you can also foresee that machine to machine communication could be provided as a private network running in parallel to the Internet and, of course, also Internet of Things could enjoy these quality of service quarkt tur provided as a specialized service in parallel to the Internet access.

So thereby the conclusion is that these three separate service categories foreseen in the 5G should be compatible with the European Net Neutrality regulation.  We have also had discussions in Norway with the different operators and Barack also at the European level has had discussions in workshops and public consultations where we have explicitly requested for specific use cases in the 5G that should, could be a problem under the European regulation, and we are still waiting for examples, concrete examples that are not compatible.

These are not provided so far, so based on my general presentation, it explains in the broader lines how this is foreseen to be compatible with Net Neutrality.  So 5G is highly welcome, of course, and regulators are paving the way for this so to facilitate achievement of 5G development in the different countries that is also a goal we have as a regulator.  Thank you very much.

>> MODERATOR:  Thank you very much, Frode Sorensen.  And actually this is a very good segue to have then Bob speaking about the technicalities of 5G.  Just one second, before you start, Bob, I think it's also very good to add another element of clarification.  We are all speaking about 5G as if it was something very easy to make happen, as if it was a simple software update.  It is not.  5G actually demands enormous network investment in fiber, in micro and small cells spectrum options and that is the reason why real 5G so far exists only in China, South Korea and a little bit ‑‑ please, I don't want to steel your time.

>> BOB FRANKSTON: Since I have a short amount of time, I will skip ahead to the main point which is the primary purpose of 5G is to kill the Internet by making claims that, it's about money and economics and the use cases doesn't make sense.  It's only cellular.  Land connection stuff, the stories don't fit together.  So let me start the green button to move ahead.  So the first thing we have to understand is what is the Internet.

There is no physical thing called the Internet.  The Internet is a technique we use to repurpose any available technology in order to basically provide connectivity between end points.  So it started out by buying teleco wires and repurposing those, and it's the technique.  So it doesn't make sense to compete with the Internet.  In order to provide these services, you have to take facilities off the table to make them unavailable.

And you have to claim we need pipes or, and I will get to the pipe story in a little while.  The real important thing, and this was very hard to understand, is the strict decoupling between what we do in the applications and the packets we exchange.  And that means economically there is no way to pay for each individual packet, which is why Voice Over IP was such a surprise.  Because there was no economic way to pay for that capacity.  It happened by accident and was discovered when there was enough capacity to do the Web, and that generated as a byproduct, it even surprised the people doing Voice Over IP that it worked over the general Internet.

So it goes against the economic model and the stories.  Now, I have got a bad habit.  This is an American thing about you are not supposed to look a gift horse in the mouth because it's a gift.  So 5G claims to be a gift.  I have a bad habit of looking at it in detail and asking does it make sense.  So the real purpose is to claw the value back into the network.  So voice calls are made on these devices.  It's just an app now.  Now, in 18670, you needed ‑‑ 1870 you needed a special network to create end points because they were dumb and you needed a reserve pipe between the end points in order to keep the signal intact.  That is a design point of telecommunications because it's like railroads, you try to sell voice services as a service, and you pay for the infrastructure as a cost centre.  Even when it went digital, the first staple in new it technology is emulation.  So the network of the 1970s was designed to get a reserve path from two points in order to guarantee low latency.  We can prove scientifically without any doubt whatsoever that you needed that in order to do voice.

There was no question we need a special network for voice.  This is why Voice Over IP was such a shock because you don't need low latency for that.  The market just worked its way around, and voice worked well without any problems in the network.  Not only did voice work well, it worked better because in order to provide the fixed pipe they needed to fix the capacity to make the guarantees and charge for it.  By using commodity bits, we have to do voice at no additional cost.  Not only that, we are able to do video at no additional cost.

That's the amazing thing because for almost half a century the phone company tries to make a business of video and failed because they had to charge extra for the video capacity, and make a promise.  Will once we accepted the possibility that video might fail and that we didn't need to guarantee low latency network, video became casual where we are zooming connected to this, other people in the world at no extra cost.  That's amazing.  5G is an attempt to deal with that problem because the problem is there is no revenue going back to the phone company, to the network to pay for the infrastructure if they are not getting any of the value created in applications.

So they have to make up a story about how you need to bring that pipe back, low latency.  And what application needs low latency, remote video reality.  Now, how many people think there is a national, international crisis, the most important thing we have to do is spend billions of dollars to do remote gaming?  That's the best they can do?  And only if you do it without a wire.  And only if you do it with the right transmitter.  You can't extend it.  The can't do wire.  The story doesn't make sense.  The other story is IoT.  Connect all devices.  That doesn't make sense either because I can get a WiFi and connect as many device.  If I own the radios, I can get as many devices as I want connected.

If you hire a really bad engineer they will say bring it all back to the centre and send it out again.  It's like you want to visit your neighbor, you got to go downtown.  The idea of fixed pipes creates scarcity because you have to take the pass off the table.  How many remember the busy signal.  You call somebody, you get a busy signal.  They want to bring back the busy signal which is an artifact of having to reserve capacity and people have died because they couldn't make a cellular call and they couldn't get a lower quality because they had to get the network service.

We have lions in the Serengeti who are tracked because they can afford a cellular account but the cows cannot o so you can't track the cows.  If the farmers own the radio as local facility, there would be no problem.  So we have to evolve beyond the business model of telecommunications because it's no longer viable.  If we create applications in here we no longer pay for the network when we make a voice call.  So look at the rote analogy, we pay for road, your local city will pay for the local roads, you pay for the local taxes.  When you get, make a route using a mapping program, none of the money for the route goes to the road owner.

So we have a business model of having roads as infrastructure paid for by local communities, and applications like driving and mapping that basically are independent of that.  They use the infrastructure.  Like when you go to the grocery, you don't have to pay every time the road owner every time you pay the road.  It's a common shared facile.  That's a very powerful model and it creates abundance.  So in the 5G model you have reserved pipes for these applications.  It's like having reserved lane on the road.

Before you get on the road, you have to reserve a lane.  If you can tolerate traffic jam you still can get huge capacity on the roads.  So there is an alternative to telecommunications and alternative to 5G which is I call the national packet infrastructure.  If you just provide open packet connectivity everywhere, suddenly you get abundant capacity, the IoT things work, all of these applications start working.  Yes, you might not be able to do remote surgery, which is also called murder, because the problem with the depending on the network for things like low latency is you are fatally dependent in this case literally fatally dependent because if something goes wrong with the network your application fails.

If you are doing an application, you would be a fool to depend on 5G, but to only work with all of the 5G transmitters, which by the way, are very problematic transmitters, very limited range.  If it doesn't, if you are in the basement, you are not connected.  It's a bad design.  I could understand why engineers are excited about the 5G radios.  Like in the 1960s the sugar companies paid engineers to say fat was buy.  Basically the engineering is used to support the story because low latency on one radio requires a complete different business model to extend to the network, but having a packet infrastructure means you can get connected devices to work.

One problem with 5G for connected devices, you need a big relationship for every device, and that means the fatal point of failure.  If you had national packet infrastructure and a way of paying for the infrastructure as a common good like roads, then you could freely connect to the device without a problem.  So 5G is basically an attempt to bring back the 1970's network of failed idea because it basically is economics.  We have alternatives which generate basically abundant capacity and prosperity and are inherently neutral because packets are indifferent to the content and inheart neutrality is better than trying to oppose neutrality with hypercomplex rules that can be gained.

Sorry to speak so fast, butter out of time.

>> MODERATOR:  I guess there are comments or questions about this last point on 5G.  Is there any comments in the room other wise, I have one.  In line of what I was mentioning before, I think there is one, has though distinguish between marketing and reality.  So, again, the fact that in many countries 5G is not really anymore marketed as 5G, but 4.5G, which is a simple upgrade of LTE technology, which actually when it is in the latest evolution has more or less the same latency and bandwidth capacity as 5G may offer for consumers.  This means that actually 5G is very, the reality of 5G is very far from the marketing.

So one may start, therefore, to wonder if 5G concretely may not be used as a, let's say, a Troy horse to ask for concessions from regulators and for a more flexible regulatory environment rather than doing the work as operators in investing in network and not trying to create new business models which actually are obstructionist.

So my question is for the, my August panelists, but also for the room, if you have in your countries ever found a, not an operator, but a device producer, an IoT system producer or a private sector representative which is not a telecom operator asking to regulators to have special 5G provisions or telling them that the current regulatory framework is not good and they cannot develop their sewn 5G networks.

So my question is besides operators asking for upgrades or special provisions to allow for 5G from regulators, is there on the other hand any IoT system producer, any connected device system producer, any virtual reality service provider that is asking for special 5G compliant provisions?  I see silence.

>> BOB FRANKSTON:  The printers like it because they produce a lot of labels that say 5G.

>> KS PARK:  I think silence is the point that there are really no killer apps that have justified the demand for such bandwidth yet, but just I think we are going out of time, but just my intervention is 5G should be separate from network slicing.  What we were, I mean, what we talked about negatively is network slicing where you slice the bandwidth and sell it to slice at a higher price.  There you are discriminating packets, you are setting conditions for delivering packets, but 5G is possible without network slicing.

Now, whether, without being able to sell those highly priced slices to the killer app providers, whether it makes economic sense to be able to build out the network without selling those premium slices, I don't know, because I'm not an economist.  But theoretically 5G can exist without network slicing.

>> Bob:  What do you mean by 5G?

>> KS PARK:  Well, using higher frequency electromagnetic waves 20 times higher.

>> MODERATOR:  Let me restore discipline for one second.  Frode Sorensen was in line to speak and provide the last word because I see that we have to leave the room.  So I would ask the regulator to wrap up the debate and provide the rest of the comments on 5G.

>> FRODE SORENSEN:  I will try.  I think we actually had two different discussions about 5G in Net Neutrality in the panel.  We had the intervention from Bob frankston which was very interesting, of course,, and your intervention was more about do we need 5G, does it provide anything else than Internet access?  And that is one side or one side of the discussion.  The other side from the operators, we hear from them some other day, I guess, they could say wee really need 5G it's innovative and provides services in different ways than the Internet.

That is for the future to prove who is right in that discussion.  I don't take a side in that discussion.  The other discussion we had in the panel presented by the French and the Norwegian regulators is more the practical discussion.  Does Net Neutrality regulation provide any obstacle to 5G development, and that was what I tried to present in my intervention, that Net Neutrality in Europe at least and in many other regulations is compatible with 5G, and then it's for the market to show what kind of services we will have on the 5G network in the future, in the near future hopefully.

>> MODERATOR:  Thank you very much, Frode Sorensen, and actually this optimist view that Net Neutrality is indeed compatible with 5G, but I think more work is up to us to do, perhaps from here until next year's session, we may also start some specific work on 5G and Net Neutrality which seems to be something quite intriguing and interesting to be studied.  Now, let me thank everyone here, both the panelists and the participants for the very good input and excellent discussion.  Of course, more discussion will be more needed and helpful, so, yes, I have already mentioned the map, Chris, but I will mention it again.

So do not forget to check the zero‑rating.info website where there is also not only a lot of interesting information about zero‑rating, but also a map organized by country and by couldn't nexts now of zero‑rating applications, regulations, and Net Neutrality regulations.

Thank you very much and see you next year.

(Applause).

(Concluded at 1305).