IGF 2020 - Day 4 - DC on Data and Trust

The following are the outputs of the real-time captioning taken during the virtual Fifteenth Annual Meeting of the Internet Governance Forum (IGF), from 2 to 17 November 2020. 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. 

***

 

>> EMILY TAYLOR:  Good morning, good evening.  This is the coalition on data and trust.  If that's the intention you are trying to join, you're in the right place.  Very pleased to see your attendance today.  We will get started at our scheduled time in two minutes.  And continue is very good to see you in the crowd today.  I understand you have very many demands on your time.  Very delighted you've been able to join today.  We'll get started in 2 minutes.  Thank you.

Okay.  I think we'll get started.  Welcome, everything, and welcome to IGF 2020.  I want to do a few housekeeping things before we get into the substantive discussions.  Of course, for those of you who have been to IGF, we don't have the opportunity to all be together, neither do we have the tune to have the full round table experience where no voice is privileged and every voice is welcomed and able to participate.  We're in a constrained environment and I think we should all Thank the members of the IGF and the secretariat for doing such a tremendous job to get the IGF online fully this year.  We're in a sort of webinar format in Zoom.  If you, marked as an attendee rather than a panelist, we're not able to give you the microphone or camera on this occasion.  Oh, actually, Enriet is saying we can elevate all participants to panel I wants if we want more inclusion.  Can I ask the remote moderators to elevate our attendees so that everybody can join the conversation?  That would be most appreciated.

Okay.  My name is Emily Taylor.  I am the CEO of operation labs.  One of the sessions along with Giovanni Seppia, if you would like to take the floor after we run through the initial panel presentations, please either raise your hand using the raise hand ‑‑ I don't know if we have the raised hand actually or type in the Q&A.  I think this is better.  Type in the Q&A if you would like to take the floor and then we can arrange to unmute you, I think.  We'll figure it out.

So back in 2006, I was privileged to be part of the first manage in the IGF under the brilliant chairmanship.  And that's where the idea of Dynamic Coalitions first took hold.  There is a lot in that concept.  It is both in the grouping of people.  They can come in and out according to the sent, also the themes and the label of the coalition for the willing who all want to gather around particular issues and explore them in the safe space that the IGF provides for discussion, exploration of issues without the strain and stresses of particular outcomes being held or the negotiation that is there in many environments.  What happened we wanted to do this year was to bring together people to explore issues around data and trust.  We have deliberately kept that to be a very open and very wide scope.  Data, if you're in the domain industry as many of us are here today, data would immediately take your mind to registration data, to accuracy, to issues of D, in S abuse arising from poor data and the impact that has on trust.  If you are coming from more of a platform environment, more of, you know, somebody like Lisa Maria from the Oxford Internet here, you would be thinking data has information.  The relationship between data points and bringing those together to form into information and also the impact that has on trust.  What is poor information, what is truthful, disinformation.  So what we wanted to do was to bring these two communities together, the domain name industry operating below the level of the application of content, but yet very connected with it.  And also the platforms which have been studied almost as an Internet in their own right whereas if they are all part of the internet that we are supporting in the Internet Governance forum.

So that's really what we had in mind and Giovanni will introduce these ideas and explore them further.  Let me just introduce our speakers who are going to kick off these themes.  First of all, I'll call on Giovanni Seppia, who is the external relations manager at EURid he's held since 2007.  He coordinates a team of 20 all over Europe who manage their relations on registrars and registrants in marketing relations and also liaison with the European institutions.  Prior to that, Giovanni was the general manager at center.  He was also involved for many years at the Italian registry and also at ICANN as well.  His background is with the masters in EU policies and programs all Venice International University.  So we're also hoping to have Pablo Bello who is a Chilean economists and has had a background working with the Chilean government from the late '90s to 2010.  He's working currently with what's app and has also had a number of roles including on the manage and also on the global on Internet Governance.

Lisa Maria Neudert a core researcher in the propaganda project and has been exploring all aspects of public and private governance surrounding this information whether that's to do with elections, electoral interference or this year very much focusing on the so called info demic.

Antoan is a Ph.D. students studying with innovation and competitiveness.  He's joining us today as a member of EURid's youth committee and he's very well coming in that role.  But he also is a man of parts having innovative founded three companies in the fields of innovation and information technologies and has chaired several non‑profit organizations and is a former representative of all students in the biggest university in Bulgaria.  Welcome today, Antoan.

And we have Roberto ‑‑ Alberto Rabbakhin.  He tackles online misinformation.  So highly relevant to today's discussion and prior to joining the commission, Alberto is a research scientist at the Massachusetts institute of technology.

We also expecting floor points to be made by Peter van Roster of center who is the general manager at Center.  Very much welcomed today and also I'm hoping that Gran Bunton will say a few words, but I'm aware it's extremely early for participants from North America.  So that's the introductions.  Delighted to have you all here today for this inaugural session.  We're going to be talking about the impact of the pandemic and how that brings aspects of the DNS and all the platforms and I hope we can have a trillion active discussion.  With that, let me call on Giovanni Seppia and also focus on the pandemic.

>> GIOVANNI SEPPIA:  Thank you for the nice introduction.  With your support, we have been joining IGF through many forms and this year with all the pandemic, we thought it was a good student to start a Dynamic Coalition which is two areas that are not so much covered by the Dynamic Coalition after assessing what are the Dynamic Coalition function.

So the optative of this Dynamic Coalition is indeed, to explore the difficult balance between data and trust.  From multi‑stakeholder perspective so not only registries as EURid, which is in Latin and (inaudible),  but also from the perspective of other stakeholders including youth and institutional perspective.  If you do not mind, I would like to open this session with a tribute to somebody who passed away in the last 24 hours.  It was Madeline Kate.  Madeline has been an active member of the IGF from the very beginning and she passed away and, you know, as I said, it's a difficult balance between that and thruster.  I remember that Marilyn was one of the strongest people I ever met when it comes to freedom of expression.  So many years ago during a public forum when the ICANN public forums were really live debates, she had a really strong and very vibrant exchange of views with Vincent who of the chair of the ICANN Board.  We're talking about 15 years ago.  And then I was working for ICANN in data.  I approached Marilyn.  I said, Marilyn, don't you think you are a bit too harsh on vicinity?  And she said, Giovanni, freedom of expression comes first.  And this is something important to acknowledge and understand that indeed, you know, freedom of expression is as important as making sure that there's a good balance between data checks and trust.  And results another sentence of one of the latest IGF.  She has participated in.  I am quoting this sentence from Marilyn intervention.  She said the skills needed to be discerning consumer of information sources on the Internet are not taught at any level in the education system regardless of the country you are living in.  And this is so much true because again, nobody is telling us how to let's say discern, assess, ponder the information that we are literally bond from the internet and this is, you know, about data.  So all the data we have received on a daily BASIS from different sources, from different online sources.  So, you know, just a tribute to somebody was contributed a lot to the IGF and to the old Internet dialogue.  So thanks a lot to Marilyn.

With that said, the registry manager of the Latin, Greek and (inaudible) when the old pandemic started, we were upon the European commission.  When the old pandemic started, there were great concerns about the data that may have been circulated about the pandemic and also may have been associated to the main names.  In our case, the (inaudible) domain names.  We have started a process at EURid to check the quality of the data.  The process led us to prevent the delegation of about a bit over today.  5,000 domain names.  EURid has 3.6 million domain names registered and out of the several hundred thousand of domain names that have been registered in April, we delayed about 5,000 because those 5,000 contained key words that are and were associated to the pandemic.  It doesn't mean that those key words, those domain words were associated with abusers, but however, it was appropriate to check the data of the registrant and therefore, you know, so we did.  And we have to say that the quantity of domain names we have reported to law enforcement authorities that again possibly are associated to abusers and are also associated to the current pandemic is really minimal.  So we have seen that out of all domain names, only 0.71% of those that we have reported to law enforcement authorities are associated to possible abusers as well as to the pandemic.

We have also started a great dialogue with all our stakeholders to see how we can improve the quality of the data in our database and also to make sure that the data is seen as the domain name of a trust working environment.  We believe there are some take aways of the past months.  Those take aways are ‑‑ say take aways that are common whenever you start any assessment or any investigation or any action.  The first one being that the ecosystem of the DNS should be seen as a sort of family ecosystem.  It's back to work as in a family with a great level of cooperation among all the stakeholders involved.  So it's not just a registry operator job, but it is also a register operator and going back to Marilyn and educating also the end user.  The registrant and consumer to make sure they are aware that there is the need data quality for their own safety, not because somebody likes to chase people until they give confirmation of the data, but just because, you know, it's their own safety that comes first and because that comes first, they need to provide reliable data just in case something happens and you need to contact them.

And also another important take away of what we have been doing with all the checks and data is that there should be the definition of clear goals.  So let's say the effectiveness of any measure should be always linked to clear goals.  Otherwise you end up really fishing in an ocean without, you know, having in mind how many fish you like to fish and therefore because of that, you end up doing a small action which is not so much an effect on the entire system.  As I said, there's the need for greater cooperation and what I said we have had a great dialogue with our registered community to improve what we are doing.  I believe a great dialogue in data and trust environment is needed to make sure that in the future, all the actors of the Internet DNS ecosystem they work together to insure the quality of data, but at the same time, by insuring the quality of data, they also increased the trust in our environment.

So with that said, I don't want to take any longer.  I know there are other speakers.  I'm happy to answer any questions later on.  Thank you for the opportunity and thank you all for joining this Dynamic Coalition inaugural meeting.  Thank you.

>> EMILY TAYLOR:  Thank you also for that wonderful tribute to Marilyn Kate who was, you know, such a figure in this environment and a friend to so many of us.  Thank you for, um, for mentioning her and bringing her into your comments today.  Very interesting efforts being made by EURid and for some other registries to try to spark domain names linked to possible misinformation.  I am going to come back to you in the chat on those percentages and also I think the, you know, the levels of cooperation and efforts to create data quality.  So thank you very much for those introductory remarks.  Pablo Bello is our next speaker from What's App.  I was just checking to see if you have audio and video.  If not, we can skip.

>> PABLO BELLO:  Yes.  I'm here.

>> EMILY TAYLOR:  Very warm welcome to you, Pablo.  Thank you for joining and to discuss COVID‑19 trends in the social media environment and particularly from your view point in your ole in What's App in Latin America.  The floor is yours, Pablo and welcome.

>> PABLO BELLO:  Emily other for me at a pleasure to participate in this panel with all of you and be more involved in the Dynamic Coalition for me would be awesome.  Thank you very much for the invitation.  As a member, it's very ‑‑ I am very proud to participate in this new IGF in this we are complex year that we are facing now.  Very happy to be with you.

Well, the issues that we are debating today are quite important, of course.  And deeply linked within the pandemic.  Very much underneath the challenge we are facing are not new started many times before COVID and, of course, even many times before Internet at all.  We are facing a crisis in trust that is affecting the core institution of our societies and nobody can say that the current situation of the world has been caused by Internet or the platform at all.  Nevertheless, this information is a critical issue.  It's very important.  It has always assisted.  It is nothing new, but, of course, in the super connected world, this information could play faster and farther and this is a really important problem.  It's important to mention also that we still don't know how misinformation really affects our perception of reality and how in the real world how implications has been in our lives.  Some time ago, an article said Eco chambers are killing the democracy, but after that, this particular form is four years ago.  So after that many research have found that connected people are actually receiving information for more diverse sources of news outlets than the analog people.  So this Eco chamber and misinformation issues are a problem, of course.  But it is not totally clear how they have affected our real life.  I think there's a state of some sociological or psycho logic research that is needed to understand completely how misinformation acts in a different way and this is still inconclusive.  Nevertheless, we read this information to have a problem to all of us for citizens, for Civil Society governments, for communities and, of course, for platforms.  It's totally important and this is what we are doing is critical to fight against that.  So let me talk a little about COVID, the pandemic this year, this complex year for all of us.  What are we doing to fight against misinformation in What's Inn.  The first one is nothing new.  But lockdowns and quarantines have changed the way people international.  This is our new normal of business is connected and seated in front of a screen.  The Internet has been more important now than even before.  So to be connected is not the topic of our plan, but still to close the divide is pretty important and even more important because of that.  And these are platforms and what some in particular have an infrastructure to be in touch with friends and relatives to work for edge, to contact health services.  Some data that I want to share I think is amazing.  What's app at the beginning of the year surpass the number of 2 billion users worldwide.  And at the beginning of the pandemic, What's App surpassed and this is our new normal.  100 billion message by day.  100 billion message by day.  I think this is huge.  This is 5 times the peek of the S system has in the whole history of the SMS platform solution.  So it's huge.  And in many parts of the world, What's app has been a small and medium business in regards to the lockdown.  But still, 40% of the world population is isn't connected.  In many places of the world what's app and the platform are the old way that people can interact and to participate in the e‑Commerce and to buy goods and services and so on.

So COVID has accelerate the word economy and this, of course, makes our current debate about trust even more important than error.  I think the question that I think is the purpose of this ‑‑ actually the Dynamic Coalition is how can we restore in the institutions in the governments and our neighbors is not only a little problem.  It's a real problem that it is facing many, many parts of the world.  We will leave ‑‑ and What's App will address the first time the integrity and security of many communications.  And I don't want to spend too much time talking about it, but of course the foot print is everywhere.  We are doing so much more online.  So there is so much more we have to protect.  And this is What's app are making such a big push to protect the strong and to end corruption in front of policy makers around the world.  When you encrypt message end to end, it means only the center and the user can see message (inaudible) and we are facing pressure from some governments to develop back doors and balance promises regarding how the technology works.  From our point of view, there is no such theme as a back door just for good guys.  If there is a back door, hackers and hostile governments will find ways to explode it.  Working on encryption will work people at risk and we believe encryption is linked with trust and debate that we are facing today, but we need (inaudible) between privacy and security.  We lose private and we can have encryption in the name of security, we lose both.  We need to continue working in technical to be more effective in crimes without putting at risk one of the privacy and security which are both critical components on how trust is built.  So my first point is it's very important in order to protect trusts, in order to be ‑‑ to allow us to live in this world we are living on.  Security, safety, privacy are core components and encryption is a relevant part of that.

So over that, misinformation of core is one of our big most important challenge.  The word health organization called the info demic.  How are we dealing with this info demic.  Four points I want to arise today.  The first one is the most powerful tool to fight against misinformation is to have trustable channels of accurate information.  So one purity to us is to offer these channels inside the platform.  The second one is how to improve the way of the platform work to minimize the risk and behaviors linked with misinformation.  Third, how to offer more privacy controls to the users in order to allow everyone asked to be more protected to misinformation that is spreading in the platform.  Andrin crease awareness and create more tools to manage to allow users to handle, to manage this information.  Coming to this four points.  Regarding the first one, I really believe that the most important tool to fight this information is to offer users accurate information to have this channel with official data especially in a held related issues of the COVID pandemic.  It is pretty important to have channels in an easy way to allow people to receive the right information against ‑‑ about the pandemic.  So this is why what's app worked with the world held organization to set up a chat bot that we launched in April this year.  A very easy to use chat bot on What's app on the main language of the world in all the official unit that makes the language which a lot of information about the more basic questions and more complex questions about the pandemic for the user.  At the same time, What's app partner held over hundreds of national and international governments and illustration launching this chat box with automatic size information that people can navigate on.  This was very important especially at the beginning of the pandemic.  At the beginning, Mass app received more than 120 million message a week on this particular channels.  And actually, some authorities in many parts of the world started to innovate using this chat box.  For example, Buenos Aires, they create a chat bot on what's app.  The world held organization, but then they added an option to create an interaction between medical held professional with the citizens to ask the results on that and even to start some treatment.  They implemented a 3 edge on the chat and the impact of this chat bot using the three was huge.  Especially in the content of the pandemic because, of course, the pressure over the health infrastructure was pretty high and they received and they will be able to manage more than 75,000 COVID related consultations over the chat bot in the first month.  But also and this is not 100% health related, but it's a critical issue in our time and that quest works during the pandemic.  Domestic violence has been a very problematic situation worldwide.  We work with the Chilean government and the Brazilian governments to develop channel on what's app that is secure and private.  It's protected by encryption and silent to allow women in danger to Connect with the authorities using what's app.  This was one of my most important projects for me this year.  I think ‑‑ the impact was huge.

Coming back ‑‑ I will finish very, very shortly.  I think ‑‑ yeah.  In the platform itself, I think this is important.  What's app introduced a lot of adjustments and feature to reduce the rally.  This is something that I want to mention because in April of this year, what's app reviews a number of messages that someone resent.  If a message is shared more than five times, the app itself reduced the ability to share this message to just one person per time.  Doing this change alone is the reviews in 70%.  The highly forward message.  So this contributes to reduce verality and of course it contributes to fight misinformation and finally, I think it is also important in terms of providing better information, operate information and partnership with checkers and traditional mainstream media in order to offer more tools, more channels, better information to citizens.  For us to protect the integrity of the private integration is critical and very important.  This is why we are fighting inscription and introducing a lot of new features and creating a partnership with different entities with fact checkers to offer better information to seek a sense especially regarding the pandemic.  Yeah.  This is what we are doing and ‑‑ this is what we are doing and very happy to be with you and if there are questions at the end, I am very happy to answer it.

>> EMILY TAYLOR:  Thank you for setting the scene and looking with What's inn.  I can see your remarks are getting the chat very active and I hope that we will be able to hear from some of those people in a little while.  I'd like to come back to you on closed groups and how the contrast between what's app is sort of closed and encrypted environment compared to the more public environments and how you sparked verality within that what the outside world has a close the environment.  I think that your remarks lead very well into our next speaker, Lisa Marie Neudert.  Lisa, you have been looking at all aspects of the COVID‑19 pandemic building on your units work in relation to electoral interference.  Thank you for attending today for your support to this Dynamic Coalition.  The floor is yours and please introduce your remarks.  Thank you.

>> LISA‑MARIE NEUDERT:  Thank you very much for an interesting start to this panel.  I will contribute a little bit on research about COVID‑19 using public data from social media and networks and I have also brought some slides with me today, which I will share right now.  And I think that works.

So basically I am start of a project at the Oxford Institute.  We have been doing that since about 2014 and obviously got a lot of traction with our work in 2016 and more recently as Emily already said in the introduction.  We have focused on the spread of misinformation on COVID‑19 because obviously there has been a whole lot and I'll talk about some of the work that we did today.

So we are coming from, we are firm believers that misinformation is not just for an interference.  It is not just any sort of hostile actors that are during with that intention that we very much also have different culprits and algorithms and a big data enabled tools that are optimized for junk news and this information because it is so engaging.  We have alternative media outlets that are doing very well over these social media algorithms and that are also having economic incentives.  And, of course, and I think the last today has shown that again.  We really have a mainstream problem with this information.  Some have (inaudible) as politics of post truth, but we definitely see political leadership both in authoritarian regimes, but also in democracies disseminating all sorts of conspiratorial or preceding information.

So why COVID‑19?  I think by now the phrase info demic is widely known.  It was phrased by Antonio Gotieres saying COVID‑19 is a pandemic, but we also have an additional later which is info demic of misleading information that is spreading on social media where we really have a lot of citizens seeking news and information and using the Internet and social media for it.

So our project at the DOI has been doing a bunch of different things.  One of them was we have weekly briefings on Corona virus misinformation.  We have that from April to August.  We wanted to see what is the level of factual information, what is the level of junk we're seeing and also what is the level of state backed outlets reporting about COVID‑19.  So a stateback outlet is a rush out today, but we also covered outlets from China, from Iran, from Turkey where we do know they are funded by the government and they are certainly propaganda tools.  So please do look that up.  If you are interested in that, it's on our (inaudible) page.  There is also Ms. Information on YouTube.  We have a work paper on that and we also have a commentary paper on that.  Very interesting.  We also looked at how platforms are first enforcing their own terms and in terms of conditions.  So our platforms have taken content down and how long does it take them to take that content down if it's legal, if it's against terms of service and then the paper will speak more about today is your paper that I have authored with Emily Taylor which is on optimizing for an info demic.  What are web and search engine strategies that are making COVID‑19 this information so visible.

So going into this, there is a lot of misinformation that we're seeing on social media, but what about search and what about the web?  So search engine optimization is something that's been around for a very, very long time.  It has not been studied as much as Facebooks, the Twitters when it comes to the spread of misinformation.  The question was:  How can we see search engine optimization being used by professional news outlets and then by junk use and misinformation pages on the other hand?  And then what impact do third party links have on the reach of junk news sources?  I will explain that in a little bit.  And how are the major online advertising platforms helping to finance junk news sites on COVID‑19?  What we did is we looked at 830 domains that were actively publishing news and information on COVID‑19.  We had a sample of professional use and information and we had a sample of junk news and misinformation and how have we identified is we were drawing from sample that our team has labeled during the U.S. election in 2016 and also during the mid‑terms in 2018 and again during the European parliament election in 2019 and then we also screened all sorts of external fact checkers which were reporting on this information, conversation theories and having to do with COVID‑19.  We also used a couple of tools to look into content headers and commercial search engine called Semrush.  So the first helpful stack is domain authority.  So domain authority basically is an indicator for how likely you are to see a certain page when you are trying to access it through a search engine like for example, Google and so we're seeing here on the left‑hand, we have a sample of performing best and the junk news and misinformation on the other side.  What you are seeing here is a lot of the junk use and misinform outlets were really getting excellent scores.  So for search engine optimization and authority above 60 and know it's a very good authority to have, which means you are very visible on the web.  And then something else that we thought was really shocking is that here, you know, these are the top 20 or so and we see an outlet like political or also German or Deutsche which has the same kind of domain authority as Russia today.  So that is really ‑‑ yeah.  Making it very tangible if you're using search, you are as likely to come across a political as Russia today when you are seeking information.

Next we also looked at back links.  So back links is something that is also going into search engine optimization and basically is a measure for how many pages are linking to your page and the more pages you have linking to your page, the better.  And even better if these pages that are linking to you are high authority sources.  So we did look into that and you can see here the number of backlinks that junk news and misinformation pages had.  80% of our junk estimates have more than 1 million back links and more than 30 had over 10 million back links.  So that is a tremendous amount of back links which means they are very visible and have a lot of pages that are referring to them and that (inaudible) means that it is easy to find these pages when you are seeking information.  We also looked a little bit into who are the pages that are linking to these junk news and this information domains.  You can see the top two referring sites listed here, which already (inaudible) was linking a lot for two to top ten sources.  We did another thing, which is we were looking into what are universities doing and, um, also what are pages like the White House and what we found is that a lot of universities including Oxford by the way and we're linking to those pages.  We've linking them as a credible source, which is the worse thing to happen, but then also for pointing to them and saying this is ‑‑ this is, you know, R, it got comments or (inaudible), which is a terrible page that's been reporting this misinformation and we're still putting this link here.  But in turn it is more visible especially if you have a University of Oxford, if you have a Harvard or a Georgetown because you are linking to those pages and giving more authority and credibility to these junk information domains.

And the last thing we did is we looked a little bit into mark up and advertising.  So mark up basically is how is a site sort of formatted to make it more legible and to make it easier to pull out information for Google, for Facebook, for Twitter.  So that for example is a link review.  It can be how a picture is formatted.  We saw professional use and junk nail is optimizing for social media.  No surprise here.  Facebook was the most important mark up to optimize for.  Here we saw junk use optimizing slightly more than professional news.  I think no surprise given the state of the Internet in 2020.  And we also looked a little bit into advertising platforms that had space on profession use and junk news pages.  Here again, very wide sprint both among professional and junk use and this information pages, but it is still again very interesting to see that, of course, you have Google ads.  You have Amazon ads directly funding this information pages.  It is something that I think many inner community have pointed out as a problem that we don't quite know what to do about except for black listing and that is also coming with a lot of problems.  It's still, of course, very interesting to see that data behind that.

So to sum up, we have a high domain site that means they are highly visible and they are easy to discover.  They actually display evidence of professional search engine optimization.  We have major academic and major government sites that are back linking to junk news sources often unwittingly and boosting their reputation and we have an underlying layer of mainstream industry advertising that is funding junk use this misinformation on COVID‑19.  So that was it for me.  Looking forward to your questions on this later.  I more or less captured it.

>> EMILY TAYLOR:  Great remarks and presentation.  I can see that it stimulated a lot of interest in the chat.  So hopefully we can get the interventions of some of the people responding to it as we move into the discussion.  Thank you very much for that.  I think it would be really interesting to come back not only to you, but to the other speakers for the sort of mapping across to what the domain name industry is doing.  Looking at the domains in the right‑hand column that is spreading in misinformation about COVID, not one of them uses the strings that we would be looking for.  But it might be interesting for us to come back to those differences and similarities and also, of course, the very important point about that you made in the end.  The junk use sites are really keying into the advertising ecosystem helping to drive traffic threw search and also from activities through advertising.  Thank you very much for that and let's come back to those issues and, of course, the questions being raised in the chat.  Antoan Shoharov from youth committee.  Great to have your participation here in the inaugural meeting of this Dynamic Coalition.  The floor is yours for your remarks.  Thank you very joining us and please go ahead.  Thank you.

>> ANTOAN SHOHAROV:  Thank you.  I would like to share my screen.  I had to see and can you grant me access?  Okay.  It is granted.  I believe that you can see my screen right now.  First, thank you for the opportunity to participate here with all of you.  I thank you for your attendance.  This is my first participation in IGF event.  So thank you.

The opportunity to participate in this event and to have this meeting today showed a great possibility that Internet provides.  As we all know, the great possibilities comes with high risk and in 2020, the world has faced the COVID‑19 pandemic which caused dramatic things in all fields.  But if there is anything positive in 2020, I believe that this is the fostering effect of COVID‑19 innovation.  And this also digitalization.  It is estimated that ‑‑ it is estimated that 3 to 5 years of digitalization has happened in less than 9 months in 2020.  Now, more than ever, every business needs to be online in order to survive.  Also public structures needed and still need to digitalize fast to answer to the people's needs.  But through this rises danger caused by the issues that we already know like data leakage and fake news.  And in the digitalized world, leakage of data is one of the biggest dangers no matter if we speak for about personal or corporate or administrative data.  Just to give you all a sense of issue, I will say that in 2019, the economic was due to leakage of data globally.  It's estimated by IBM and (inaudible) republicans tight arranging between 1.25 and 10.19 million dollars depending on the country and the industry.  Between 2014 and 2019, the amount has raised from 3.5 to almost $4 million on average per country.  Again it is different for the different industries and different countries and companies.  And what about the fake news?  The University of Baltimore estimated annual was around $78 billion.  These figures speak for themselves.  In 2020, certain activities made this danger even greater than before.  I will share with you two very simple examples.  The first one is related to working from home.  Working from home the majority of the countries in the world because of the pandemic advised their staff and their working staff to work from home if it is possible.  But also this is the danger because working from home is not as safe as working from office.  Because of the network, because of the Internet provider, because of the whole security infrastructure that is being used in the offices.  The second examples related to the health care innovation products have been launched this year because of the pandemic or related to the pandemic.  For example, people with health conditions and (inaudible) context of global COVID‑19 related examples.  All these services collect certain private data and it is very important for this data to be safely managed and used only for certain activities for which the different user has agreed.  Just imagine if this is breached by hackers and they will be able to know your health condition and also know your location which may endanger your life.  Another major danger in 2020 is related to fake news related to COVID‑19 articles which are spread worldwide.  The great danger comes because many people believe in them and this is the result not only negative or economic results, but also on live (inaudible).

Honestly speaking, we can speak many more about danger, but I would like to add and talk about a solution.  I believe that solution to these dangers come based on joint efforts by the stakeholders because these dangers cannot fight individual.  We need to fight them as a team.  And fighting fake news can be done through reestablishing the trust in the banking institutions of the democratic society, but understand this is not a simple solution.  It takes time and needs joint effort.  Great example of this is what Pablo said earlier about the joint cooperation and those of many governments.  This is how we can fight together.  And the issue about data leakage, the only way to be to fight it again is to joint effort and preparedness.  So think about ‑‑ thanks for the opportunity and thank you for discussions and questions.

>> EMILY TAYLOR:  Antoan thank you for joint exercising and this is a simple problem to solve.  But you make an excellent point which really builds on Pablo's remarks about how the pandemic has also just create hasten digitation in a remarkable way.  For those of us who have been around for a while, we have been predicting the growth of interactions based on video convincing and yet we have been a bit puzzled about why that hasn't taken off.  Now it really is this sort of necessity being the mother of invention and the fact that people still need to communicate, still need to have togetherness and the digital environment is there for us, but at the same time, we need to restore trust.  Thank you very much.  I'm sure we will come back to lots of themes you have raised.

What I'd like to do before calling on our final speaker Alberto is to invite anybody to take the floor.  I have Polina and Rambuntin on my list of people who would like to make a contribution.  We have put you all as panelist.  Just write in the chat if you would like to make the list and then we can have the discussion started and shared amongst all of us.  With that, Alberto, can I call on you to take the floor for your remarks?  Thank you.

>> ALBERTO RABBACHIN:  Thank you, Emily.  It is really a pleasure and I listened very carefully to the previous presentation.  All speakers had a very interesting point.  I am here to give you a little bit of an overview and in particular, um, on what the unit that I work and following very closely, which is the practice on this information.  I think I agree with what has been said before.  Internet and social media provides greater opportunity also for making a democratic process more participatory and inclusive.  We are all aware of how this new technology can be misused to disseminate this information on a scale and with a decision of targeted that it was unprecedented.  This information is not a new problem, but the technology has made this problem very pervasive.  We need to do something and we started some years ago the commission on the finding (inaudible) to try to limit the impact of this phenomenal.  We are all aware that this phenomenal will not go away and we will need to work and continuously work because the technology's evolving.  So the policy needs to adapt to the technology.  Evolution.

So one of the signature initiatives originating out of the commission in 2018 communication taking the line of this information was the code of practice on this information that was after say laborious work, but very compressed work that involved several stakeholders lead to this code of practice this was signed by major online platforms.  So we're talking about Google, Twitter, Facebook and now we have also Microsoft TikTok and other platforms involved and also the advertising industry that was involved in the process and signed the code.  So the coated was let's say ‑‑ the coat was to let's say design to limit vulnerabilities of certain online technology and services throughout five pillars within this code.  And they look to limit the availability of website, increase transparency of political ads, and often involving this information campaign and power users to make more to make them more resilient to this information and the power research in order to better understand the phenomena of this information.  So these are the five pillars of the code.  The code was signed and implemented in October 2018 and in two months ago, we published the assessment after one (inaudible) of the code.  It is done on the work between October 2019 and September 2011.  It took a while to make this assessment because it was ‑‑ we counted on a lot of contribution.  First of all, the self‑assessment of the platform.  We then have the report of the European regulators group for all your visual media which is maybe better known under (inaudible) and a study done by independent consultant and the commission report on the 2019 election to the European parliament.  So it was a very reach information that we have used over to generate a piece of this assessment.  Of course, the code is a unique example of self‑regulation on this information and applies in Europe at the moment.  Other parts of the world have interest on having the code implemented in their region.  And we have to acknowledge that the code has been an important and necessary first step to create a more transparent online ecosystem.  I think one of the issues that reduced the thrust on the digital media is a lack of transparency.  And certainly this code has improved on that (inaudible) because we have now a framework for a structure dialogue between the relevant industry actors, the commission, the national authorities.  Through this framework, we are better informed on the platform policies against this information within the European union.  So this is I would say a general positive achievement.  (inaudible) substantive progress.  The resources are substantive progress that's been made on the implementation of the code.  I don't want to go over a long list, but certainly most of you might be aware of the fact that in view of the European election, all the major platform created at the libraries where they included the political (inaudible) that were say used during the electoral campaign and then this library became also valuable for national elections.  So this was certainly an important achievement although we are well aware of this repository present limitation that needs to be counted.  Another improvement was regarding the integrity of services.  We saw that the platform took action against accounts using manipulative techniques to artificially amplify the reach and the impact of false and misleading information.  Also they started to provide more information, more detail under efforts for these accounts and they use (inaudible) and they also improve their communication regarding the action they have taken to limit coordinated information campaign by (inaudible).  So also under this point of view, we have data transparency of what they do, which is of course important also to understand their impact on freedom of expression, of course.

Empowerment of consumers they certainly improve and make investment to give Providence to more ‑‑ trust forward information sources.  I think here the user has been set by Pablo, not only What's app, but all the other platform have created spaces where they promote information coming from public health authorities including the world health organization and professional media also.  So this is something that they achieved during the crisis and was not achieved before.  So we were really pleased to see the prompt reaction during the crisis.  They also ‑‑ we have to acknowledge also have increased whole duration with a fact checkers in many new countries and in particular, this also happened during pandemic and the (inaudible) during the pandemic.  We have commissioners issue communication.  In June related to the COVID‑19 misinformation and one of the results of this communication is that we are now receiving monthly reports from the signatories of the code of practice where they detail what they have done to take online of this information.  So this is again another improvement from the transparency point of view.

Of course, these are the positive side, but also signature for short comings.  And in particular, there is inconsistent and incomplete application of the code and platform and (inaudible).  So certain tools, for instance, for the citizens, for the user not available in the space and are not available in all languages.  So this is something that should be (inaudible).  We didn't see that much effort in limiting the online ads to appear on website that prevail this information.  I think what Lisa showed before is a clear example.  We see how the ads goes more into websites.  Say nor inclined to have low quality information than to websites that have higher quality information or information that is created according to professional and ethical standards.  So even if you have the list, it is quite long but I don't want to go too much into detail.  I want to emphasize an outstanding problem which is the fact that access to platforms data for purposes that remain a major concern.  There are only bilateral agreement between platform and university to access limited sets of data.  This is not enough to make the researchers and the conditions to better stab it.  The information phenomena on all platforms because we know that on certain platforms, it is easier than on other platforms.  Also the crisis, the info demic linked to the crisis shows the importance of better ‑‑ uniform definition of certain behaviors or certain let's say issues that are linked to this information because if we have ‑‑ we are having a better set of uniform definition will help also define a more (inaudible) set of policies to take over the issues that are at stake.  So that's very important to progress on that.

>> EMILY TAYLOR:  Thank you, Alberto.  I'm sorry.  Is it possible to wrap up your remarks?

>> ALBERTO:  I was coming to an end.

>> EMILY TAYLOR:  I thought you were, but I thought I would mention.

>> ALBERTO:  (inaudible) and in particular this area that is the lack of keeper format indicator and which is are missing.  So it is very difficult to assess in an independent way and by civil sort as well to what has been done by the platform if we are not able to define this and we don't update to measure the indicators.  And, of course, the (inaudible) approach so far has been limited to the signal which includes the bigger platform, but not all the platforms.  So this has created also an Asymmetry between those (inaudible) of the code and those that don't have a signature of the code.  So I'm done.

>> EMILY TAYLOR:  Thank you so much and emphasizing the code of practice.  And also, you know, highlighting action, positive actions taken, but also the imbalances between the large platforms where we have had all of the spotlight and perhaps other environments which Lisa Marie has covered.

What I would like to do now, we have 15 mentor so left and what I would like to do is to call on first of all Graham Bunton and Polina Malaia just to make a brief intervention for a couple of machines, but there are also many more who have been putting interesting remarks in the chat.  Maybe Megan Richards, Lori Schulman, Olivier Miguel.  If we can try for the 15 minutes if we have time, wrap back through the panelist of just very fine remarks.  It's great to have all of the chat.  So Graham, can I start with you?  Thank you for joining us.  I know it's early and I will go to Polina and maybe to Myra.  Thank you.

>> GRAEME:  I'll try and keep this brief and share a little bit about our experiences as a registrar in COVID and some of the issues we were having.  So my name is Graeme.  Continue is primarily wholesale and we sell to other companies and not by and large the general public.

So we recognized relatively early on in the pandemic that this was pretty exceptional circumstance and that we needed to be a bit more engaged proactive and tried different approaches to respond appropriately.  So by and large as a registrar, but not a hosting company, we don't typically interact with the domains on our platform to look at them to see what content is on them that's really outside our purview and typically outside the appropriate course of action for us.  So, you know, pandemic hits.  This seems different and we should maybe take a closer look at what is going on.  We set up some queries and key words and began looking at the registrations since they were coming in every day.  This peeked in late march for us, but we were seeing something close to 300 registrations a day.  That is by the end of July, we have 17,000 total and as of today, something around 22,000 related, COVID related registrations, which is a lot.  So we were trying to assess them for harm because this is again such an exceptional thing.  So we were look will at them by hand, a team of people every day clicking on those demans to see if there's anything there.  By and large, 70% were just part.  They weren't doing anything.  There's a variety of stores selling EPE and blog websites with some information.  And the stuff we found was harmful or potentially harmful.  Say fake tests, fake cures.  For us harmful really presented an eminent threat to human harm was more than half percent.  It was a very small amount.  In trying to set the demands, we had to look at making this process more efficient.  There were lists being circulated around the DNS community.  They were by and large terrible for the most part.  They included ‑‑ there was not a lot of robust activity in creating them.  So people were putting all registrations in with abusive registrations and then someone else would repackage that and add some more on to it.  So this list ‑‑ here's all the bad domains including things like officials in the south African page and things like hospitals and health care providers.  And so we were in a really tricky spot where we didn't have any reliable sources of information to act on.  So really did have to go through this exceptionally slow and manual process.

>> EMILY TAYLOR:  Thank you.  One more thing?

>> GRAEME:  Those were often not at all related to the strings we were looking for.  So we had no mechanism to find them.  They were using sub‑domains.  If I'm thinking about what we can do about this sort of thing, it's better feeds and better data.  Better quality, more information and more evidence and then really the thorny bits of this for us were outside the world of the DNS.  Here's a blog.  It's got four posts and sort of generic COVID information and one that is clearly dangerous.  Who assesses if this is harmful and it really felt inappropriate that it was us as a DNS operator.  And know there was enough of that and no one sort of within our community and law enforcement or governments really felt like they were in a place to deal with those sorts of issues.  That was a real challenge for us.  Thanks, Emily.

>> EMILY TAYLOR:  Thank you.  You highlighted the limits, but we were all part of the same ecosystem.  The different ways in which people are using these domain names sometimes harmful, sometimes not.  In fact very little and the importance of data quality on the lists circulating for people who want to take action to help.  Polina, thank you very much, Graeme.  Polina, can I ask you to take the floor?  After that, can I ask Myra to take the floor.  Very brief comments if you can, everyone, so we can fit in as many remarks as possible.

>> POLINA:  Thank you, Emily, very much and thank you for this opportunity to speak up today.  I will try to be as brief as possible and just quickly I am Polina.  We greatly welcome this initiative and I would just like to also briefly highlight on the point about data.  Also Graeme spoke about and it is permanently highlighted in the name of this coalition.  Primarily we feel there is a need to discuss the issue of inconsistent reporting of the scale of the problem within the so‑called (inaudible) discussions.  And also specifically during the ongoing COVID‑19 pandemic.  So the research that center conducted at the start of the pandemic is fairly consistent with also the figures that ICANN has reported and also what Graeme talked about before.  Only a few dozen domain names among the tens of thousands that got reported by some security companies were problematic.  It could be associated with some type of abuse.  And these domain names do pose a real danger, there's no doubt.  However, I need to measure the problem so that we can identify the best way to solve it.  We see that the low amount of domain names associated we have identified as part of our research.  This number is actually manageable with the regular processes that already are in place across the domain name industry.  And specifically namely with close corporation that comes with authorities which is law enforcement and judiciary processes, et cetera.

Lastly, we agree on the data sets that help support the public policy objectives and increase trust.  So thank you very much.

>> EMILY TAYLOR:  Thank you so much, Polina.  We very much wanted it to be a form in which they could be, you know, can generate evidence based for policymakers to help combat harms, but also not over reach into legitimate activities.  If I may ‑‑

>> Sorry, Emily.  I think you muted yourself.

>> EMILY TAYLOR:  Thank you.  I did mute myself.  I wanted to give you the floor to either contribute to the discussion or raise a question.

>> Thank you.  So I worked as a public policy officer for the Australian government and I was leading the governments response on tackling the sharing of crisis attack that happened last year on social media platform.  We introduced a legislation that gave the Esafety commissioner powers to regulate and direct platforms to take down that sort of information.  But this was based on purely reporting method.  We then started working on tackling this information in relation to COVID‑19.  Surprisingly it came from the toilet paper issue that was in Australia.  So now my question to the panel is:  In building the relationship between data and trust, which the panel mentioned, is it just a digital problem, but a real problem, but often with COVID‑19, what we're seeing is Ms. Information from online is being shared amongst family members and neighbors on what's app platform.  How do we collaborate and tackle that because what's inn is a very closed off platform.

>> EMILY TAYLOR:  Thank you very much for that.  I will take a couple more interventions and hopefully we'll get that answered.  A really key point there.  Thank you for that.  Can I call on you, Megan Richards to raise your questions or highlight some of the issues you have been doing in the chat.  Thank you.

>> MEGAN RICHARDS:  Thanks, Emily.  Very interesting discussion.  It is similar to what Humayra said and many others.  Packs are facts.  We sometimes have alternative facts.  I was wondering if there's a technical or legal or statistical way that one can identify back these facts out by foot noting.  There must be some ways and perhaps some of the participants have ideas of how one can show that this is really factually reinforced information that is being provided which will then, of course, lead to more trusts.  So that was really the question that I have.

>> EMILY TAYLOR:  Thank you very much.  May I call on you Lori?  You have been very active and Miguel (inaudible) and I'll go back for some very final quick interventions from our panel.  Thank you.

>> LORI SCHULMAN:  Hi.  It's Lori Schulman.  A few thoughts come to mind.  Of course the values.  I was very struck by the list of the fake news versus the non‑fake news.  The left and the right columns not even labeling them fake and non‑fake because there's huge spots of people and we have seen it this week with the U.S. elections.  News sources that have been viewed as trustworthy and non‑trustworthy.  Names we saw on the left.  I think it could be problematic to draw such great lines T. goes to Megan's point.  How do we verify?  How do we tell the world this is real when reality or the truth is looked at as something relative rather than absolute?  I think it's a fundamental question.  It's a question we see in jurist prudence.  Now we see it every day in our lives on social media.  Why do we put a value system against judging a fact number 1.  Pack facts are facts and now we're living in a world with (inaudible) facts.  How we verify is key doing it in an automated way.  A lot of people have talked about ways in the domain system of verifying ownership and we than can clap back against ‑‑ but (inaudible) to some of the concerns people have for individual privacy rights, but at the same time if we don't hold individuals accountable, how do we get to that truth element.  Again, being in the field that I'm in had in terms of trust and the trust crisis, one of the things here although we're talking about COVID, something that is clearly not something subject to intellectual property right or discussion, this does go to a broader issue of how do we identify trust, what trust is?  I do feel compelled given my advocacy to say that when the property rights are discussed in different context throughout IGF and any platform where we've talking about rights, it is very important that rights as rights sometimes are forgotten that they emanate from points of trust.

>> EMILY TAYLOR:  Sorry to cut you off.  Everybody who has spoken today I can just ‑‑ I think we can all talk for hours about this because people are making such relevant and thoughtful interventions.  Really thank you so much for that.  I would just like to ask Miguel to take the floor if there is time and if you're able to and then I'm sorry we're going to probably have to wrap up.  I will try and put the question to Pablo if we have time.  Thank you.

>> MIGUEL:  Thank you so much.  My question is a quick question.  Do we have a lack of competition in several domains of Internet (inaudible) networks or what's app and then we need more competition.  We need to be more InterConnected.  We need to Connect the phone networks and intranet works and now we have socio network we need intersected in a way users can interpret user (inaudible) for example.  Specific question for Pablo.

>> EMILY TAYLOR:  Pablo, you have a couple of questions about interconnection, but also I want to diagram back to HUMAYRA's point about sharing within family members on What's app.  Pablo, can we have your very brief remarks before we close the session.

>> PABLO:  First, we believe for privacy and security, but because of that, it's not possible to moderate content in an encrypted platform.  This is a key issue for that.  You can take down content on what's app platform.  This is what we can do and what are we doing.  First, reduce verality by design and reduce a lot of friction to reduce verality.  Second, combat messages because there are some providers spamming on what's app is something against our times of services.  We combat that.  Third, offer channels for information and fact checking I think the collaboration with fact checks are very, very important.  We have organizations and governments providing accurate information.  And finally, I think at the end, the most important are users responsibility.  So this very important to have a broader conversation and to teach actually users in how to be responsible using the tools and the platform of the Internet and technologies.  This is a key issue.  Regarding interoperability, it's not possible to interoperate platform that has different technologies and actually telegram is not created by design.  So it's a technical issue that is not easy to fix.

>> EMILY TAYLOR:  Thank you very much, Pablo.  I would like to come back to all panelists and I would also like to call on many others in the chat.  Unfortunately, we are totally out of time.  But this has been such a wonderful start to this conversation and I hope that you will all join the Dynamic Coalition mailing list.  I hope we can continue our conversation on that.  Very much thank you to our wonderful speakers today and to all of the participants who made such great and thoughtful comments either on the microphone or in the chat.  Thank you very much, everybody.  Have a great rest of your day or evening and let's keep in touch.  I think there's a lot we can do together.  Thank you.