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.
>> WADZI MOTSI-KHATAI: Good morning, everyone. If you could, take your seats. We're about to begin.
So good morning. Welcome to the open forum of the Internet Governance Forum. Thank you for joining us today as we discuss the future of artificial intelligence and sustainable development.
My name is Wadzi Motsi‑Khatai. I will be your moderator for this discussion.
As we know, artificial intelligence presents the new frontier for the industrial revolution. Able to leapfrog industrial revolution. This field comes with capital, capacity, access to information as they're not universally available in emerging economies. Today we'll focus our discussion on a new initiative called the Fair Forward Artificial Intelligence For All initiative. It aims to facilitate more inclusive and sustainable, and, like its name, fair practice in artificial intelligence.
Sharing their perspectives on this topic today are our guests, please help me in welcoming our guests to the panel today. If you could, give them a round of applause.
( Applause )
>> MODERATOR: We will begin with Mr. Holtsburger. With over 20 years of experience, he's worked for policy for the green party. Today, he'll be introducing the project to us and speaking from that perspective. You have the floor.
>> Thank you very much. Good morning, everybody. Distinguished participants of the IGF.
Ms. Romanoff, good morning.
Wonderful, ladies and gentlemen. My name is Mark Holtsburger. I'm happy to welcome you on this open forum of the IGF.
This panel today is called the Future of Artificial Intelligence and Sustainable Development. It's hosted by the United Nations global paths, Ms. Romanoff.
It's following the preview on Monday on AI in Africa.
In the following 60 minutes, we'll, at this meeting, unveil a new ministry on AI.
Let me briefly tell you a story from four weeks ago. As you know, Germany has connotations with its partner countries to discuss the priorities of how to achieve social and economic progress. Recently, we had these discussions with India. For the first time, AI was the topic ‑‑ the top topic on the list and how we could collaborate on this issue.
At the end, Germany and India have shown to have learned one lesson, that digital transformation is not optional.
As a member of the division, can I tell you that my ministry focuses on many topics of up most important, fighting poverty, ending hunger, or addressing the effects of climate change. However, we know we need to rely and to build on the potential of digital approaches. If we want to achieve the SDGs by 2030.
Developing and emerging countries cannot and should not be excluded from this ongoing fundamental technological change. This is how we interpret the "leave no one behind" principle of 2030. Otherwise, we're face a gap and growing digital divide. As of now, half of the world's population is offline. Fewer women are online and less than 1% of all persons worldwide derived from least‑developing countries.
So what is my ministry doing? There's a clear strategy on digital transformation. Second, they've stepped up the area of digital transformation and is running over 550 projects in over 90 countries. Third, we now start to focus on key technologies for development. AI is, without any doubt, one of these key technologies. As I said, we need that digital technologies and therefore, also, AI to achieve the above‑mentioned SDGs.
On the other hand, we're convinced that artificial intelligence has to be anchored locally, and we say AI should not just address the benefits of a few but should, instead, empower all people and communities.
From this background, I am proud to announce the official kickstart of our flagship project FAIR Forward artificial intelligence for all.
( Applause )
>> I'm thankful and really honored that we're doing this just not on our own. For goodness sake not. We can work with partners in Asia and Africa and worldwide. Some of them, we're tell you in a second.
But coming back and to explain the approach of our flagship project, let me just explain to you the abbreviation FAIR. You all know this word, FAIR is F‑A‑I‑R. Let us start with the F, like framework. FAIR Forward advocates base AI rooted in human rights, international norms, and privacy. FAIR Forward to support our partners to develop specific policy framework which will allow a locally rooted, responsible, and ethical use of AI.
Ms. Romanoff will tell us something about it in a few seconds.
Let's move forward to the A, like anchor. FAIR Forward will anchor its technology in Africa and Asia.
I, we will drive to integrate digital training and learning for the development of AI in our partner countries.
And, finally, we have the R, like responsibility or responsible data. Many of you will know that knowledge of AI programming is not enough to create AI applications. In order to develop and train AI, you need data ‑‑ and I mean a lot of data. But one of the major problems in many places are missing data or biased data.
In fact, many of our partner countries face a lack of accessible data that fits to the local context. And that is why we want to remove empty barriers to AI and improve access to local training data and open‑source AI technology.
Let me give you just a brief idea of what FAIR Forward is aiming for. I don't know if you all know ‑‑ sorry. I do not know if you all are aware but all voice systems work on AI, and all of the major commercial applications understand English, French, Chinese, and even German, but none of them understand any African language.
They work for the sake of a few and not for the majority of the world. Try, for instance, to ask Siri how weather will be in Uganda today and tomorrow. That regularly drives Siri mad.
We believe open access to African and Asian voice data is a key area. That's why our flagship FAIR Forward wants to support developing technologies so more people can benefit from this digital transformation.
I'm looking forward to you, Mr. Niyonkuru. We'll talk about his ongoing work there.
Delivering to populations in their own language, in order to include interaction with government services or to provide small farmers with voice‑enabled services.
There Lukas Borkowski will give us his vision of such a future.
Let me talk about our partnership with Mozilla. We not only work on the collection of voice data. In fact, we want to build an alliance for open voice technology were African‑Asia. I recommend to check Twitter for more news on this partnership. All the activities are based on a close dialogue with our partner countries.
This is why we're here today. We want to start a dialogue with the IGF community on how we can make sure AI supports sustainable development in the future, and how we can avoid the growing digital divide.
I will end by repeating our vision for FAIR Forward. Our aim is to build a more open, more inclusive, and more sustainable approach to AI. Let's build our forces on. This I look forward to our discussion with you and to the panel now. I want to thank everyone here for joining online or offline to bring development corporation closer to digital transformation.
( Applause )
>> MODERATOR: Thank you. Thank you very much for that inspiring opening.
Our next speaker is Audace Niyonkuru, CEO, Digital Umuganda. As you've just heard, digital Umuganda is developing digital voice assistance in Uganda. Here to share his perspectives on AI and sustainable development, please join me in welcoming Mr. Audace Niyonkuru. You have the floor.
>> Thank you. I will be talking more about the R in fair, which is responsible data. As you know, we've been working with Mozilla to build open‑voice datasets. I'm going to cover how and why we're doing that.
You might have seen this number over and over again at the IGF. 53.6%, this is the official number from the ITU talk about the number of people in the world connected to the Internet. 53.6%. You can easily say that 53% of the world is connected to the Internet, which is a good milestone. But the opposite is not connected. As you see, much of the unconnected people are coming from developing countries, mainly 28.2% in Africa and 48.4% in Asia‑Pacific countries.
The Internet is heavily skewed toward English and other major languages, and having open‑voice datasets will help solve that challenge. English is only spoken by 20% globally and 5% natively. You can imagine what this means for the 43% of the people that are not connected to the Internet, not only because of lack of infrastructure but lack of access to services and information in the local languages. This is mainly the major challenge that we're solving.
The question becomes: How do you make technology more inclusive for people who do not speak major languages that are accessible on the Internet?
We believe through the partnership of Mozilla and others, open‑voice dataset is going to be the solving solution to that.
That's why we partner with Mozilla as well as others on bringing this voice datasets. We're doing that through the community. So it's not only about the way we access data but how we build it, how we maintain it, and how we govern it. We're doing that towards a more community‑based approach where the communities participating in building the voice datasets but also the innovators in the community will be the ones who are innovating on top of these datasets.
Here, we have a system of voice technology that would otherwise not be accessible because of the heavy investment needed to bring the infrastructure needed for voice interaction in local languages.
And we're doing that. As you see, our name is called Umuganda. If you been to (?) A lot of people have been there already in the crowd? Quite a few. If you've been to Kigali, every Saturday, we have people coming together to build infrastructure. This is roads, hospitals, schools, depending on the needs of the community.
How do we bring that to the digital age? Because with the digital age, new infrastructure has to be built. We're building a solution that's rooted in the culture to build different infrastructure because it engages 80% of the population. It can be adapted to this infrastructure as well. So it came from people working on normal infrastructure like this and has been upgraded to this.
This is one of our Digital Umuganda events and this is how we are building the dataset, by having people come around like this to build the dataset. Not only are we working with universities, not only are they helping in building it but, also, they're the innovators that will be building solutions on top of it. So in some sort of way, we're also leveling the playing field so not only global companies are able to build solutions but also local innovators are able to build solutions for the local communities and, hence, allowing innovation to come from all over the world in the places we would least expect it.
Thank you. Murakoze, that's one last word you can learn.
( Applause )
>> MODERATOR: Thank you. There will be time to ask our panelists questions following all of their presentations.
Our next speaker is Mila Romanoff, Lead Data Policy and Governance, UN Global Pulse. She's responsible for establishing sustainable mechanisms for private partnerships as well as responsible use of big data for global data. Her experience expands litigation and more. Join us in welcoming Mila Romanoff.
>> Thank you and thank you to our partners. I'm happy to be here to continue these discussions started on day zero and also opening up the panel speaking about challenges and opportunities. I'm happy to give you a little bit of introduction on the work we're starting to do and the work we've done so far.
So global policy is a special initiative the U.N. Secretary‑General which is tasked of bringing up the value of big data and artificial intelligence to help the achievement of sustainable development goals.
The function is three tracks. The key one is scale and discovery. The third one is policy. Policy includes understanding and also developing mechanisms for the governance of data and the governance of the artificial intelligence. I will speak about the governance today.
If we are currently working ‑‑ we have a lab in Kampala and we have our regional lab in Asia, in Jakarta in Indonesia. Through our regional hubs, we are working with the governments across the regions to, first, understand the challenges and try to try to address these challenges with the key partners. I will talk about how we're doing this in the African region.
We have started work with partners a few years ago. We'll continue this work going forward within the next two or three years. We're looking at how ‑‑ well, first of all, what value artificial intelligence brings to the Global South. And through that, we're partnering with academia through key partnerships in developing tools and developing the software that, first, helps understand what type of data do you need, what type of technology do you need to understand the needs of the local communities in the Global South.
We started this work in Ghana and Uganda. One of the interesting projects we've been working on is actually speech‑to‑text through collecting publicly available data from radio. As you would all know, in the western world, we're working with social media. Social media is an active resource.
In order to understand the localized languages, we have to build software, right, to understand the languages and then to translate it into the context of those economies.
That's one of the projects, and there's many more. The key question, when we think about it is: What type of frameworks do we need to understand and utilize this value? The framework is around responsible use of such technologies. The framework is actually not only bringing the data and artificial intelligence in the economical sense but also empower people from all sorts of human rights perspectives.
As a key here, the fundamental question here is the right to privacy. We've been working on developing the data privacy and data protection regulation with requests of different governments across the Global South as well as within the European Union. We've started work across the United Nations system as well to as well unlock the power of data through developing data governance frameworks within the U.N. specialized agencies.
Through this work, we're hoping, based on our own learning and experiences and own work within the U.N. system, we hope to translate the same experiences and pilot the local communities bringing them through the processes like in Uganda and also in Ghana, to build the governance framework and help the governments build framework for artificial intelligence and data to ensure that the value of data is unlocked.
Just a session before that that took place basically an hour ago, the special advisor to the Secretary‑General brought up an idea that was translated by the Secretary‑General of establishing the regional help desks as one of the ideas to bring the local communities and the local needs into the development of such government frameworks. That's one of the key recommendations that came out of the report of the high‑level panel of digital incorporation.
To build frameworks around data and AI, it's the context in which data will be used. We should never forget the context, right? So one of the recommendations also of the high‑level panel on digital incorporation to the Secretary‑General was also in order to understand this, we need to bring the context, in particular the context from the Global South. So what are we trying to do in order to address these challenges? So in the work we're doing now in Uganda and Ghana, that started actually with supporting the government of Uganda in developing their data protection law, now in actually developing and supporting the president's special task force on the fourth industrial revolution to develop the ethical AI framework supported also by our key partner BMZ and this will continue into the future years as well. It's built on the key consultations from the stakeholders coming from the local communities. That's one key part in it.
What was mentioned, also, in the previous discussion is data. Data is the key element, right? So before we start with the governance frameworks, we also need to understand how do we govern data and how do we unlock the data to feed into the artificial intelligence. So the second part of our work will concentrate as well as in Ghana and Uganda, in building the sustainable economical infrastructure to unlock the power of data and develop frameworks, both the regulatory frameworks, the government frameworks, as well as the technical frameworks for sustainable data access.
I want to stress that the key is sustainable. As we all know, the data now is actually being locked and held by a lot of the private sector companies around the world. This is not only the power of the Global South. So all the role here and the goal is to build the sustainable frameworks that would allow the private sectors, NGOs, to share the work with the public sector and the other way around while protecting preserving the right to privacy and also bringing the right technical solution into place. So we're starting this. This is a huge project and a quite important one. It's actually figuring out the regulatory barriers and the technological barriers. We'll be looking for the partners on both fronts. The process will start through a series of consultations that will be conducted at the national levels and also regional levels, and we'll be seeking support of the expert group on data governance and AI, the global policy as established in 2014 which also has experts on the rotational basis from around the world.
In addition to the expertise of this group, we're, as I said, conduct local consultation, bringing the experts from the community. At the IGF, I hope we find that expertise, and that expertise will be reaching out to us within the next month and years to contribute to these really crucial discussions.
I would like to thank you again for the time and the invitation to be here. I will be happy to answer any follow‑up questions. Thank you.
( Applause )
>> MODERATOR: Thank you.
Our final speaker is Mr. Lukas Borkowski, European Partnerships Lead & Country Director Madagascar, Viamo. They help individuals make better decisions about technology use and development. Mr. Lukas Borkowski is a strategy and development expert. He's here to share his insights as one of the new partners in this initiative.
>> Thank you very much. Thank you for the opportunity.
I think I brought you a picture. If not, I can also try to put your focus back on our target populations and on the people we actually do this for, the artificial intelligence and sustainable development. If you imagine a small farmer in Uganda who is most likely part of the 43% that was referred to, we have to find a way to actually give them access to information in their moments of needs, in their local language, and through the devices that they already have.
While we all hope for having universal mobile data coverage anytime soon, that's also a big dream. So tackling that opportunity and that challenge right now is what Viamo is doing. We're a global business. We operate in about 30 countries. I represent my local colleagues that actually do all the hard work. What we do is combine center design, behavior change, communication, and, in the end, digital and appropriate digital technologies, and we sit between operators and social media channels and partners or local governments to facilitate meaningful interactions with beneficiaries, meaningful interactions with the small farmer, for example, or with young mothers that have questions about the health of their children.
We try to change the paradigm in communications and give them access to communication through, for example, simple mobile and telephony. They're not only part of those with lack of access to Internet, but they're illiterate usually. They do not own smartphones typically. Given those factors, classic mobile phone communications through SMS services or other services is meaningless because they cannot read or write a full sentence and cannot interact with that communication. That's where they come in and work with our telecom partners and partners like BIZ. Right now, we achieve meaningful interactions with 19 million unique users per year in about 300 million sessions in 2018 alone. It's expanding not only because of the opportunity but because of the need. People need access to information when they need the information.
We cannot wait for them to wait for a community health worker to wait for a radio show or something else. They all want to have access to information as much as you desire access to information right now in your smartphones. So I think that is very critical. What we see with it, it's very powerful. 75% of all callers to our hotlines get set up and make it to key information on best practices in breastfeeding or how to grow maize. That leaves out 25%. 25% of people who call, the system may be too complex for people who are it literal and never went to school. That's why we're interested in voice recognition and why these partners are coming together. We want to work on voice recognition and unlock it in local languages to make these services more accessible for local populations in their local languages and to close the 25% gap and do even more.
For now, as it was mentioned rightfully, the capital costs of developing such an algorithm and the technology behind it is extensively high, even for a social business like Viamo. This is something we couldn't do ourselves. So this opportunity of making it available as a source is a game changer and a change in the paradigm in communication in international development.
If you imagine people in rural communities in Nigeria or India can call in and simply say, I want to know what the weather is going to be tomorrow, they will be able to make important decisions about growing produce in a much more effective way, not losing money anymore because of bad decision making and actually coming much closer to getting out of poverty. If young mothers can call in and say my child has blue pickles, what is the problem here, you will actually save lives with that.
I think it's important when we speak about framework and speak about regulation and responsibility data theoretics all very important, but it's really important to keep in mind why we're doing this. We're doing this to reduce poverty and the save lives. That's what we're trying to do.
I think there are a lot of ethical questions on the way. To be fair and frank, we do not have the answers to all of these. That's why work done is really important. As a private company, in the end we're a social business, but that also means a private company with subject in all of these countries to local data regulation. Sometimes it's even contradicting to our ethical understanding of data privacy.
So it's not always that clear, and there's no clear way forward right now, and that is why this work is important to really advance this and bring this in the future. I think what we can contribute as the private sector, certainly data. We have 6 million interactions right now every day on our platform and through those services. So we can feed that data back in. I encourage over private sector players because open‑source technology and open‑source voice recognition will not work. If we keep it behind doors, it will serve the very few and will only be accessible to the big multi‑nationals. That's why we follow this open‑source. I thank you for the great initiative. Let's take that forward.
( Applause )
>> MODERATOR: Once again, thank you to all of our panelists for sharing your thoughts.
We'll take two questions from the audience for our panelists to respond. Then we can take the next round, time permitting.
We have you, sir, and you, sir. Then I will come back.
>> AUDIENCE MEMBER: Good morning. Thank you for a wonderful set of presentations. I am from IT For Change. We're in India. We're part of the coalition that comes together and how we can build what works for everybody.
The issue of data protection and making data work for the Global South, I think many speakers spoke about it. I want to bring one element for discussion and for the views of the panelists.
Basically, if you look at data from a protection point of view, there are two aspects. One is privacy. You don't want to violate that. You want to protect the interest of individuals. But I think we need to look at another aspect, which is not so much talked about, which is economic value of the data itself.
Now, I just want to bring up a critical perspective ‑‑ I will keep it very brief, but this is an important aspect, and I think it's ignored. I want to quickly bring it in.
Countries in Asia and Latin America have been for centuries been a source of raw materials which they export to Europe and the developed world, and then they import back the finished goods. The terms have been so unfair, which is one of the primary reasons why developing countries are struggling, whether it's Uganda, Africa, or any country in the developing world, we can see the same issue.
If you look at today's situation, there are companies in the world, data companies, it's profitable not because of a software platform that may be free or open‑source. It's rich because it owns data. I think requesting them to share the data, that's the biggest asset and the companies clearly tell the governments, The data belongs to us.
How do we make sure the developing countries do not continue to remain exploited by the private‑sector companies and how can we make sure the digital divide is not going to cause extraordinary inequality and inequity. I think the situation is going to be much, much worse than the situation we had centuries back.
I think the issue of economic ownership of the data is very critical. Understand by mentioning that the coalition has come out and there's reports available in the lobby. Data subjects must own their own data individually. I would urge the panelists to think about and respond to.
>> AUDIENCE MEMBER: I'm an MP from TLC. I've been in different sessions, and I was reporting that we have a problem in Africa that we are now ‑‑ the big companies are getting data from us, but are not helping us with actual intelligence. It's good to know the German corporation have decided to help Africa country to catch up on that.
But my concern is that the way maybe the project was designed, maybe the phase one ‑‑ I don't know if there will be other phases ‑‑ like the case for the good project in (?) I will share the problem with that. It's the same story. You're getting help getting the data. That have, I don't know what will happen. It's somehow developed on technology. You should also help them to not only provide you the input but start even start themselves the input. They are a step before other countries in Africa.
Also on the language that you choose ‑‑ ( non‑English language ) ‑‑ few people know that language, but if you want to help Africa country, there's a language ‑‑ I think everyone here, if you haven't gone to Africa, you know hakuna matata, that's Swahili. So the impact of that project, you have also in Uganda the same project. If you can associate that things of getting from them and making the project in Kenya and Uganda, it can help more African people where we need the voice services.
We don't have people to read or write, but if they can get data services through voice, it will be much better. Thank you very much.
( Applause )
>> MODERATOR: So, are there any takers for that question? First looking at not exporting data as a raw material and considering other options for the project.
>> I think I'm going to answer the gentleman from India as well as the MP from DNRC about data being exploited by the Global North, not the Global South. I think a lot of what we're doing is working with the system and the innovators in the space to bring the solutions in the infrastructure. It's not just about having an infrastructure or dataset that nobody would use, but working with them and looking at solutions that would practically be there. It's also an equation of looking at how do we level the playing field for local innovators? Usually this data is held by companies and multi‑nationals.
We're giving ability to local innovators, and there are a couple of those that come to us and say they have the solutions. The innovators are there. What was lacking was the infrastructure.
Yeah, I believe the ecosystem is there and getting them use of this dataset is also something we're doing and will continue to do in the progress of the project.
And I guess for the scaling to other African languages, maybe BMZ would be more equipped to answer this, but I believe the recently announced BMZ partnership is to scale to other African languages as well, yeah.
>> I would like to share some of the work that we're doing to address the exploitation and also the frameworks around data to bring up its economic value. We've conducted a series of consultations over the last year. They're continuing. The next one will be in a few weeks in Uganda. So far, we've done a few consultations in Ghana in cooperation with the ministry of ICT and also the Data Protection Office. Some of the outcomes ‑‑ prior to that, we hosted in Nice. The question kept coming up about how do we bring the value without jeopardizing the human rights, particularly the right to privacy.
The second question that came up was how do we bring into context in the African economies in particularly and what can we do about that. How can we ensure that the data does not get exploited and stays there? We don't have the answer to everything, but we do have a roadmap, at least, from the way the U.N. is looking at it.
One, as mentioned earlier, is to bring up more and more experts from those countries. We can talk about the regional, but we also need to talk about the national context. Africa is big. Every culture is quite distinct. Even the communities within each country may be quite distinct, one from another. We really need to keep in place the local context.
So that's one key part to it.
And we need to make sure the people, the people from these communities, are actually the ones that are deciding and developing the laws and the frameworks and the standards. And how do we get there? This has been raised in previous sessions of IGF? We have strong regulate frameworks and examples from other parts of the world. It has been brought up, and we've been listening, but a lot of the times, the trainers are being used as examples in the Global South context.
So there's advantages and disadvantages to that, as to any question is to any solution.
The frameworks that have been developed are quite tailored to the western world economies. Taking those frameworks and bringing them into the Global South context can work, but only if they're not copy and pasted as‑is. If they're looked at only as one of the examples of the best practices. It can be shared from the U.N.'s perspective. The U.N. also has a specific ‑‑ it's all the same. It has a mandate. It has a specific status, and it also needs to operate within a specific frame. So we also look at the best practices, but we also need to keep in mind how we're functioning and think of what operations we're delivering.
So in the context of the Global South and particularly in Africa ‑‑ I will speak specifically for the African context ‑‑ we need to help. In the U.N., that's the role of the U.N., to bring in the context and the need, the social and economical needs to the context of these frameworks.
One example is, you know, the ‑‑ I'm not an African, but working with a lot of African colleagues is that notion of openness and the notion of collective rights and how this is realized through the African context, through the African culture, how can we bring that to what's shaping around the world and make these frameworks work for the African economies?
That's a question to the audience, those particular from the region. Maybe you will have an answer of how we can do that.
Third is thinking about the collective rights, right. So we look a lot about the privacy as an individual right, but more and more so, especially when we think about the Global South context and especially African context, is especially the group rights. How can we use data to benefit the group rights, not only one individual but actually the community? I think this is a distinct factor, learning how we can translate the use of data and the risk and harms it brings to the group harms.
Also, how do the group rights relate to the use of data. Thank you about it from the perspective of also using data for the good and also not using the data for the good. How can we balance the two? I think a lot of the best practices and framework that are shaping around the world really forget the factor of what do we do if the data is not used. If you have a tool ‑‑ as a humanitarian, if you have a tool in front of you and you need to save a life, would you use that tool to save a life? There's not enough regulatory frameworks and best practices recognized at the policy level that actually bring up this issue and allow for a safe and secure data sharing and use by public sectors.
And the fourth comment is also the industry. Industry is the key player here. We should not forget that also we need to think about the small businesses and all the regulations around the world. The best practices are oriented in the businesses. We need to think about the capacity, the capacity of the small businesses, the capacity of the small communities, and bring that into perspective.
Fifth point, capacity itself. Many have Laos and have initiatives projects for laws. Many have AI governance frameworks, but the question here is the capacity. Once you have these laws, once you have these frameworks, the next step is the implementation. I encourage us to think about the capacity that needs to be built within these local communities to actually implement these laws.
Which brings me to my final and sixth point. Youth and education are the crucial components, the key components, to bring this over. There will be a need for resources and a need for corporation to make sure that the experts that are in these countries understand these laws. Bring them to the civil society as well as to the government. There's a huge lack of capacity, and there's not enough of education and digital literacy brought out by these new technologies at very, very early stages of education.
So I will close here and welcome any more questions or comments. Thank you. Hopefully, that was an answer to the questions raised.
>> MODERATOR: Thank you. Maybe we can take a round of two questions. Keep them brief, and we could have two brief responses from our panelists. I know there's questions behind me and some over there.
>> AUDIENCE MEMBER: I will be brief. I read that the training dataset would be open for crowd sourcing. I wonder how do you support local communities to feed the training dataset. Thank you.
>> AUDIENCE MEMBER: I'm from Nigeria. I run an IT firm, Contemporary Consulting. We do know there's been concern regard to how far AI can go, either for good or evil. The question is: Are we incorporating accountability by design in this project? Accountability by design. It must be accountable to us because it has the possibility of getting out of hands down the line.
>> MODERATOR: Thank you for those questions.
>> Thank you. I would like to respond to the last question of how far can AI go. I give you a real concrete example. So at Viamo, we're experimenting what we call internally ‑‑ forget me for that ‑‑ the Netflix approach. It's an engine for small farmers. We all know mobile phone numbers are not unique identifiers. That's something regulators don't want to hear because we want people to sign up when they buy a SIM card. You go into any rural phone, you can buy a phone with a SIM card that's unlocked to an individual. That makes it difficult to develop recommendation engines for rural populations all around the world.
One key technology can solve that is voice biometrics. So voice biometrics is like your fingerprint, but it's your voice. It's used in global banks. It's super secure. It's only super secure if you have security in place. Security here ‑‑ I want to be frank from a private sector ‑‑ security is safe in our basement, but it's also about the ethical security of the data through regulation in those countries. We operate in quite a number of high‑risk country where is I'm not so sure if we would want to have that voice data linked ‑‑ or that voice print linked to a mobile phone number to get into the hands of various actors in that country with vested interest.
That's a critical question because AI can go really far, but do we want to take it that far? It's rather a question of accountability by design is where do we put in a stop gap. I'm not the one saying let's curtail the AI, but we have to be smart and enforce data security. We had that discussion already in a few sessions.
Just quickly because we were speaking about industry as the big player and economic value of data. I'm going to be brief on that one. When I was young and I needed the money, I studied economics. So you always have supply and demand. I think there's an oversupply of data and not a lot of demand. Specifically, not a lot of demand from the private sector. It's very true we have to educate people and work with local companies and local entrepreneurs to get there.
The biggest demand for the data that we see for now is from the government base and from political actors. When we speak about the economic value, we also need to speak about the political economic value of the data. From a private sector perspective, at least from our point of view, we need regulation, and we need work done by U.N. global policy to make sure that the data protection regulation in those countries not only protects the data but also protects the data from political actors and vested interests.
>> MODERATOR: Would you like to share?
Okay. Any further comments?
>> I think there was a question about crowd sourcing. If it's about data collection, I mean, if you go to the common voice platform, it's accessible anywhere. I would encourage you to put in the Moroccan languages as well. The way we're doing it to engage the community is we're building around the education community. So specifically around universities and high schools, we have a program called the Commoneers, students engaged in data commons, and they're encouraging other students to engage in building this common infrastructure as well as to take advantage of it and build solutions for local problems that would be around, yes.
>> MODERATOR: Thank you very much.
Thank you all for joining us and engaging in this discussion on the future of artificial intelligence and sustainable development.
Some key considerations that were highlighted today, how can we use AI to serve and support development primarily, focusing on equity access and partnership and accountability as we've heard from our audience.
If you're interested in FAIR Forward, we have representatives here. Stand up so people know who you are. If you have questions, comments, and would like to get engaged, please feel free to look at these two representatives after the panel and engage in conversation.
I would like to thank our hosts today. We look forward in continuing those discussions forward. Have a good day. Give everyone a round of applause. Thank you.
( Applause )
>> MODERATOR: It is my understanding there's a session after this. If we could exit quickly and have our conversation in the foyer. Thank you.