IGF 2019 – Day 4 – Raum I – Concluding Breakout Session: Data Governance (Part I)

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. 



>> Hello, everyone.  Thank you for coming.  We will start the session on governance.  Michael has just arrived.  It's excellent.  We'll start in five minutes with introductory remarks.  Then we'll go through the sessions.  We decided two more times for a session.  One hour sessions before reporting from the six breakout sessions and then quick conclusion to the session.

So first off, why we wanted to go over the success of the governance and we're very successful during IGF the last few days and reflecting on what we said as a speaker on Tuesday.  I guess we're on the good side of it and we have a number of suggestions trying to integrate the requirement directly into the governance.

So, I wanted to quickly come back to the six (?) we identified and saying it needs to be integrated in processes and strategies.  And know on the last state of the IGF after a number of sessions that we have cross‑border data.  We had the chance to discuss the flow of data that I can contribute innovation when we are aware to discuss it in this matter.  Also, we had interesting conversation on data and we need to stress the social agreement.  We (?) showing and using data.

We had on the track data protection framework.  So need of dialogue and all stakeholder Groups and obviously bringing to the table Civil Society which sometimes is less.

And on the force track, the issues brought forward is a matter of balance between interests and the need of cross‑border agreements.

We also discussed human right issues and the need of ethics brought to Data Governance raising a number of new issues.  For example, the need to learn critical thinking so that you can really accept and understand what data is made of, what it is used for and so on.

And lastly, we have the governance and ethics of AI and algorithm track.  We have the number of initiatives first UNESCO when we started developing a framework and human rights approach and just enter the two year process.  We have also oh CD principles,s number of initiatives, national and regional level.  For example, in France, we just started with national AI strategies and bringing over the need of common and (?) approach in the mix.

With them, Wallis, we're very happy to have a chance to lead the session and we have some main element was basically a social element adding people brought together to discuss these issues between organizers and lead us and participant at the IGF giving each other and potentially making connection and discussing issues and bringing these issues by reporting to the global IGF.

So the plan is very simple.  Pardon.  Yes.  So back to the it.  We have six groups and one hour to discuss it with every moderator and rap tour.  They will build upon human right, ethics and governance and ethics of IN in algorithms.  So just ask every group to gather introductory remarks and thanking in advance the moderator.  So bringing the reports forward.  Ben, do you want to add something?

>> BEN WALLIS:  Yeah.  You want to put the agenda back on?  I guess we had a lot of people in the room on Tuesday.  Obviously at the end of the week, some people are no longer in Berlin.  So just thinking a little bit.  It was clear on Tuesday but actually the breakout teams would have preferred more time and there was so much discussion.  There are few people here today.  So I just think we should be adaptable and we can kind of follow around the breakout groups.  If you feel you come to a natural conclusion of your discussions today, that's fine.  We don't need to take the full hour.  We don't need to go until 1 o'clock.  I will do whatever is useful.  It is an interesting opportunity for you to be able to discuss what you might have heard in sessions throughout the week that relate to these particular sub‑themes and potentially to kind of compare where insights today relate to whether it might be changes relating to the conclusions that were coming to on Tuesday.  If you went to the page on the IGS schedule for this session, the reports of all the breakout sessions are attached to that Group as a reference point.  You can have rights up and it was great to get the reports back and also write up what we get today.  I think for me what I saw is the real value is the fact people are getting together and having really interactive discussions, hearing different perspectives, maybe getting to know people who are working on the policy issues from different perspectives and different organizations and maybe hearing about sessions that will happen during this week.  I'm not too concerned about what the written output is.  I think the value is in the discussions in the room today.  So I wonder ‑‑ I noticed you had the slides with the names of the breakout leaders.  I am wondering if those generous volunteers could stand up and ‑‑ if I go one by one to help you all ‑‑ Miguel, yes.  If we go from the top, Miguel is in the back of the room.  And then we have Beto.  And on the other side of the room, (?) is in the middle ‑‑ last but not least, Maria Paz on the governance and ethics on AI and algorithms.

In case you weren't here, the way that the seats behind the table are set up, they ended up being six groups of chairs.  So it was quite natural for the breakout sessions to go to those chairs.  You can turn them around and sit in a square or a circle or whatever.  We will begin with the breakout discussions.  Thank you.

>> Thank you.

>> Strictly wise, we didn't have a conversation because suddenly, there was no one in the group.  So I had a very long conversation with myself.  And then sovereignty issues that was a very nice group to discuss.  I did want to share with you some of the ideas that were on the meetings and as well since some ideas that were reflected on different sections that took over cross border data.  What were the big issues and where lies the unrest and the hope for the future.  So basically data crossing more often and more borders than any time in the history of mankind sounds big, but it is not right now.  Now we have to reconcile the rights of the end users and the interest of companies in order to have a proper development of international standards, regulations and legislation.  And this need to be according to many, many voices that I heard throughout the week, these need to be constantly updated.  We need to sit the singing, we need to put settings and we need to be able to show very much flexibility in the way we do things in order for it to evolve with situations of the percentage in the future both near and long‑term future.  And the other particular point was the effect of decision making from companies affecting legislation from different governments, from different countries, from different territories.  Legislation wise because you have both economical and social and the same data creates resources for different countries and different companies and different jurisdictions.

So the other one ‑‑ one of the other issues was the difference of the gap now that we have lots of different kinds of gaps.  The developing gap between developed countries and developing countries that in regards of the level the quality of the regulation that they have already developed or not and how everybody else is subject to those rules even if they are not, even if it would not accept it by the parties sometimes.  They are limited to do what those regulations, foreign regulations tell them to do.  This can be develop and developing or develop and foreign regions.  How to deal with this challenge in situations.  We are in a UN setting several times corporation is international corporation is one option and the preferred one not particularly in the sense of the convention of international regulations.  More in the sense of international corporation both by regional between two regional Groups or country to country.  And, of course, taking into account the far bigger number of players such as companies and society.  The thing of the community itself in order to show education and take to the debates and processes around the big issue of Data Governance and narrow issue of how we treat data crossing borders.

As a conclusion, I wanted to put this information through you and let you know that because it is intertwined with all the Groups and you will have to speak a bit about data ‑‑ cross border data while talking about your own findings and your own Groups, seems to be that data itself is in the heart of every discussion we have and cross‑border data is going to change the world in the way we see national states and the way we see international community and we should make the most of it to achieve the 2030 agenda.

Thank you very much.

>> Thank you, Miguel.  I would like to give the floor to additional sovereignty issue.  You have the floor.

>> Thank you.  There are a lot of topics that are related to those issues and the approach that we chose was basically to ask people who were the taking out that they had from the sessions that they attended.  And in some cases, we didn't cover absolutely everything.  There's no effort of being synthetic.  It was more picking a few topics that popped up that at first didn't seem to Connect close to the issue of sovereignty.  To actually understand why there was an underlying problem that was related to that and so I would take four elements quick Lee.  The first comment that was made connected to several sessions that introduced importance ever taking into account the layers of the Internet when we talk about sovereignties and the fact that there is a strong difference when we talk about THE Internet between the logical infrastructure which is still global and not really fracturing that much.  And the application s and con tent layers that are much more subject to the diversity of national lows, et cetera.  In each case, there are exceptions because even at the infrastructure layer, there was a very interesting argument but important which is that in the case of (?), which is a land lodge country, the desire to have sovereignty and guarantee of access led the country to establish its own cables.  So there is an argument about sovereignty at the technical level, but the most of the debate is taking place at the higher level of content and, um, there was even a comment, I think, worth reporting on which is that there might be alterior motives for people who actually are purposefully or systemically conflicting the two and saying the internet is fragmenting where they don't make the distinction.  So the first message was to keep in mind the distinction between the layers, what we qualify as the governance of the Internet and governance on the Internet, but whatever the label is.  Keep in mind when we talk about the internet, there are different layers and that governance has to be adapted.  The second element started with misinformation.  At first there was a bit of balance including for me as moderator because it didn't seem to be a direct relation and actually after some discussion, it turned out that the connecting point was the notion of integrity of the electoral processes.  And fundamentally when we talk about misinformation, we can have very divergent view on whether restricting content is the appropriate strategy even if we don't agree with what people are saying, it needs to be left online because it is part of the democratic debate.  But what is clear is that the motion is a key element in the sovereignty of nations and communities and that in particular, what was of concern was the misuse of data for targeting users according to the orientation or specific qualities.  So this is how the connection between misinformation was made and I think it was an important element.  So integrity electoral process.

The third one was emerging around the notion of data sharing.  And we had an interesting discussion on the fact that even if there is the technical infrastructure, there is layer of content.  There's also another layer that is purely economical that is the economy data.  Or the data economy.  And that it functions in a slightly different way than the other economy environments that we're accustomed to including around the notion of the sharing of data.  And there was an interesting discussion about the data sharing for public interest data, all the approach of public interest sharing which was the object of one and also development purposes and what one of the participants called digital industrialization.  There was another distinction, which is interesting around data sharing which is sharing among people, among individuals and that Connects to the question of data possibility, for instance.  There's the sharing among companies where sometimes it's been official for the companies to share appropriate data because it adds to the database that can develop.  And the third element being the public interest objective of sharing of data for instance, collecting traffic data from certain apps to help functioning of smart (?)  I won't get into details, but what is very interesting is it showed the importance of governance of data, the connections between three temperatures.  Two are very familiar in the environment.  One is free flow of data and on the other hand, data protection.  There was a thinking that we should think in terms of a triangle which is data sharing or data mutualization maybe.  A did I mention it is insufficiently adopted or taken into account.  And this was a transition to the last point that I want to finish with which is that the way data functions and the economy of data calls for agreements internationally and it calls for specific shared responsibilities between the governments, the companies and so on and the last point that we discussed was there is at the moment too much debate probably on where those discussions should take place.  Should it only be among governments, WTO or other international organizations?  Should it be multi‑stakeholders and among a few actors or everybody around the table from the onset.  This is detrimental to actually addressing the topic because we can spend a lot of time discussing on what is the appropriate venue and meanwhile, we don't address something that is actually at the core of the debate.  And this is a loop with the very first point I was making because in governance of the Internet, we have institutions that work more or less.  They do work.  ICANN, IGF.  And governance on the internet, we have IGF and we have a system.  But for Data Governance, we don't really know where those discussions actually bring all the actors together.  So these are the four points that we addressed and it was actually a very fruitful discussion.  If anybody after the different presentations from the Group wants to add to it, may have forgot something.  Thank you.

>> Excellent summary.  Moving on to the next track data protection frameworks.  I would like (?) from the web foundation.  Yeah.  (?) leave us and give the floor to Stephanie which will report for data protection trap.

>> Hello, everyone.  I'm Stephanie.  I'm the mission Ambassador from Hong Kong.  We had a very fruitful organization on one of the difficulties of data protection framework we can see.  We can see that currently in the law perspective, there are difference jurisdiction in different countries that makes it very hard for the countries to really harmonize a global data protection framework.  And even we can see that different countries actually have their own needs for data protection.  So we come to a point to find out it is actually hard to find a one side face all solution in terms to establish a global data protection framework.  And also when it comes to ethical perspective or comparativeness, we realize it would be a goodness selling point for the industry.  And, um, but still it's very hard to find a common ground and maybe also indeed, many views are very contradictory when it comes to data protection.  There's also a difficulty in monitoring and evaluation and how we can ‑‑ we still have further protection for our open data and big data.  So what can we do?  So we raise an example from Kenya that currently collaborate with the counsel of Europe that they actually had a very localized data protection framework.  What we can see from this example is we can have a bottom‑up approach to establish this data protection framework.  And mean tile, it could motivate more part of the Civil Society to participate.  So, um, what's more ‑‑ yes.  So we also touched on many other issues like data deletion and data sharing.  So we had an interesting idea.  So we're saying what if we did have data literacy license like our driver license to insure everybody has literacy to use data and to use the Internet properly?  So it would be an interesting question to continue this discussion.  Yes.

>> Thank you, Stephanie.  Moving on to the data and sustainable development track.  I will give the floor to (?) from this side.

>> Thank you.  My name is WAI‑men.  So thank you to the moderator for inviting us to be ‑‑ I think it's nice closure.  We started off this and coming back to this.  No doubt we have a smaller Group and also different people, but we still have this discussion.  In our Group, we have the government and somebody from the research and from the business.  So it is really a good make and we have interesting discussion.  While looking at Data Governance for the development, you look at more like the one thing that is said very clear and data of the global self.  That came up in a few discussions and that is also (?) with some other discussions that happened this week.  It is identity they target and what we need to do.  I recall my priority data or high value data when in some instances where they have as much as 90% of the data missing, there's a need to look at priority themselves and making the best use of the resources.  I will highlight then perhaps three points on what the Group felt that there was some but not sufficient discussion during this IGF and that could perhaps be considered as for the flowing IGF and by the host countries like in Poland next year.

So the first one is the (?)  Data governance and there seems to be an under position of the importance of the data literacy that could mean how to read and write as a basic element, how to consume data, how to produce data, how to assess data and the data is also different types.  In the introductory, I recall from one government to say a lot of developing countries, there is still a lot of analog data.  But today, discussion is also a point where we have so much data.  Perhaps how do we along at (?) data to analog so you can consume by families and committees that actually literacy.

Another point missing from the discussion is on the implementation.  We discussed about the different data governance model, we talk about policies, talk about standards, protocols, we talk open data for open license data.  But as specific implementation and monitoring the implementation looking at the output, outcome impact on important paths.  We do know we are all making challenge, making reference to GDPR.  But, even for GDPR, there are many challenges in the interpretation of the implementation.  But the last step the Group or so discussed is there is also the linkage of the data governance to multi‑disciplinary approach including research.  We had a lot size policy, but there seems to be a need to look at policy in the face.  There's actually this question about what is actually needed for institutes and other academy Civil Society to look at everything and all the challenges that they get.  Thank you.

>> Thank you.  We have two left.  Internet and governance and ethics.  You two merging of the two trucks.  I want to give the floor to Maria for some rising descriptions.  Thank you.

>> Maria:  Thank you very much.  We also have (?) and Julia will presents the summary for the conversation for the two sessions of the merger.  And basically we feel that there's still a lot of energy and engagement of the participation that can know provided for the future conversation and the future (?) of the IGF.  I'm very glad to report in the general sense and I will invite also as a final remark to all the other Groups and participants to do the same.  Around this conversation when they leave this room, but rather the other way around to find a way to keep engaged even to the BBF work that is intersectional around the year using the resources ever the report that will be available in the website of the IGF and also many of the gap that were identified here in the conversation in terms of the question or the topics that were not (?) this year very good suggestions for proposing workshops and all that kind of session for next year.  So I invite you all not only from my Group, but all the Groups that have been participating in this conversation today to keep that energy and to kind of channel in the participation in the IGF.  So Julia will give us the report.  Thank you.

>> Julia:  Our themes were rather similar between the two Groups.  So we decided to merge.  The main topic is what possible frameworks can guide AI governance and regulatory approaches going forward.  We talk about what's happened and there are three main things e merges.  First ethics frameworks.  Currently there's an incredible plurality of frameworks and it's difficult to understand what is common between them.  We had a discussion of how we pull these together.  Topics such as do no harm came up, but we think there might be others agreed upon.  We also discussed how we can understand each nation's view and tools and strategies for going above that.  Some people discussed to ask different nations specifically where others pointed to a need to significant policies and strategies and they have their own AI strategies in place.  We discussed how this is a challenge and even agreeing on a definition for AI is incredibly challenging.  Many groups are using a fluid and growing definition.  They don't want to allow this to have a discussion about policy at all; however, we're hopeful at UNESCO is currently in the process of making a human rights based AI framework.  And this will be the first piece of standard coming from the UN body.  Second topic in terms of regulation, woo had a discussion of how people believe AI framework is toothless and they're as ‑‑ AI has a tendency to take them from the past and move them in the future.  This means that specific countries should be able to regulate where they want and do not want ax I to be used; however, other members mentioned how this is important to see who is setting policy and to have a real worldview in terms of cultural context and people trust in their governments.  In terms of topics for next year, measures hope to see more and more discussion of the morals of AI.  How far do we want to take AI in terms of protecting and what makes us human and AI as a supporting role.  And a discussion in job displacement and E‑governance.  Thank you.

>> Thank you, Julia.  I would like now to open the floor for comments, remarks and for the description.  We have five minutes and then we'll set the next tabs and (?)  Thank you.  You have the floor.

>> Thank you very much.  My name is Hos Kramer.  I live in Germany.  What I suggest to give the whole procedure a little bit more management oriented structure like getting openings structures of governance, the term of governance means that they discuss strategy and initiate Group to work on these policies, procedures, processes, controls and implementations to manage information together with not only the current techniques and messes, but those techniques and messes need and said operational requirements.  Those are sub‑structures of governance rather than solutions.  So that is what I'm suggesting to do to make clear what we understand on governance.  And then we see what kind of discussion groups are missing.  When I see the problem of really coming up with the governance procedure, then one of my suggestions would be easily have topic oriented governance topics oriented and not (?) topics, but governance oriented sub‑structures in IGF regional conferences and come away from (?) communities as they gather here to seize the benefit of top down structures of professional societies like one of the structures will be the international counsel of science with not only the science counsel, but its members down to the national organizations that come up with sub‑structures.  Not only us top people, but those people and address them makes them aware this is really important and we need their opinion.  That would be a broader thing and also financially asks research foundations not only offer to come here, but asks the research foundations and technical portions to really make a position statement rather than if you're not able to take place here.  I think I will leave it like that.  There are some other things that cross borders and will be necessary, something of the practical thing.  What I didn't see discussed is self‑destroying data where some kind of finishing line or kind of restrictions of data that we ‑‑ what to do and what not to do with that data.  Not just put it on the tame, but also to say what not to do.  We don't have implementation means of this.  Everyone needs it.  Something like that.  Thank you.

>> Thank you on the left.  You have the floor.

>> Okay, yeah.  What I saw this year was there has been (?) spoke about it.  In earlier years around data, a lot of talk of free flow of data and data protection was is there and data sharing and that brings to the table the aspects was brought up.  And that also brings not only personal data.  We are obsessed, but they work largely with personal data and non‑personal.  AI is the one value and how it will be distributed.  The gee on politics has a big economic aspect to it and non‑personal data aspect.  This started coming and I hope in the next IGF having spoken this time, the sharing was as unity.  How would it take place and what are the institutions of sharing of data.  It is not something that sharing will take place.  And how would that take precedent is very,very important for develop think countries and digital industrialization.  Thank you.


>> Thank you.  Last comment from Beto and I give the floor.

>> I am extremely happy about this format.  I was honestly cautious initially.  Although, I've always advocated for tracks in Internet governance forum.  This is the first time there is this book ending of sessions.  The short discussion was really, really substantive.  The notion that people have an opportunity to piggyback and reflect on what their experience in the different meetings is really great and I strongly encourage that this is ‑‑ we conducted in the future IGFs because it is really a good exercise and as a moderator, I really, really enjoyed it.  The second thing is as I mentioned at the end, one of the main benefits of the IGF is that we set aside where the discussion should take place to focus on the substance.  And this is something that is really missing.  It's a bit of self‑promotion because it is what we tried to do in the Internet and policy network.  We should be more open to discussing the substance and one typical example is what (?) was mentioning around data sharing.  Yes this is something that Ipersonally take out of this IGF is there were a session I could not attend because of a conflict, but we discussed it a little bit here.  I think I mentioned how does the data economy function and what are the governance things that are related?  How do we optimize economic and social value at the same time.  Is the real governance issue a real topic and I do not care where this is being discussed, but it needs to be discussed.  So thanks for having provided this opportunities.

>> Thank you.  That's really encouraging to get your feedback on how you felt this format went.  I know the IGF secretariat at the end of every IGF does stop taking exercise and they call for comments from the community about how they thought the year's meeting went and how it was organized.  This was ‑‑ my chair is encouraged to be experimental and think of new ideas and sometimes the night is cautious, but this is something new this year.  So those MAG members who were involved in the sessions whether it's moderators or breakout moderators or participating were continuing to next year and can reflect back when we start discussing IGF 2020.  I agree.  It seems, you know, you can ‑‑ everyone can go along to workshops and listen.  If you're lucky, you might get to ask a question at the end, but this provides a place where people can have substantive discussions and everyone can participate.  I think that in itself is valuable.  The reports were hopefully succinct and I'm going to just collect them and publish them as a report.  I also note that this afternoon 3:00 p.m., we begin with kind of the wrapping up in plenary.  So there will be sessions where each of the thematic strings have to pull back.  I have 3 minutes and I can't promise I will be able to incorporate everything that was discussed here.  Between 3:00 and 4:00 p.m., we'll hear back from the thematic streams as well as things like the high level panel and best practice storms.  So that might be a bit more of a challenge to here about each governance.  I wanted to thank you all very much for your contributions and for making it a success.  We will take back (?) the view that shared and this worked out.  Thank you very much.