The future of Digital Public Infrastructure for environmental sustainability

Exploring the convergence of digital public infrastructure and environmental sustainability at UN-led event

An event on the future of digital public infrastructure (DPI) and environmental sustainability, moderated by David Jensen from the UN Environment Programme’s Digital Transformation Team, brought together stakeholders from both the digital and sustainability sectors. The event aimed to explore the intersection of DPI and environmental sustainability, and to facilitate a dialogue on how digital transformation can support environmental goals.

Robert Opp, the Chief Digital Officer from UNDP, provided a keynote presentation, explaining DPI as the digital equivalent of physical infrastructure, such as roads and bridges, which is foundational for economic activities. He emphasized the importance of DPI being interoperable, accessible, governed with safeguards, and fostering innovation across sectors. Opp highlighted the potential for countries to build DPI using open-source digital public goods, with examples from India and Estonia. He discussed the role of DPI in facilitating environmental sustainability and green transitions at scale, such as by enabling payments for environmental services and improving data transparency for emissions.

Ana F. Vergara and Kurt Stockinger presented a new UNEP report that underscores the need to connect environmental data with economic and policy information to develop effective environmental strategies. The report advocates for DPI to enable data exchanges, improve data availability, and reduce barriers to data sharing. It also highlights the importance of bridging the conversation between different domains of knowledge to leverage DPI for environmental sustainability.

The panel discussion addressed the challenges of generating political will for climate action and the role of DPI in making environmental data more accessible and actionable. Yolanda Martinez from ITU discussed GovStack as an implementation layer for DPI, providing technical specifications and tooling for digital services. She highlighted the importance of data-driven policymaking and the need for capacity building, including e-learning, to enhance government digitalisation efforts.

Questions from the audience focused on the integration of sustainability into GovStack, the strategy to speed up DPI implementation, and the issue of political will for climate action. The panelists agreed that demonstrating the business case for climate action can help generate the necessary political support. Sally Radwan from UNEP suggested that creating the right incentive structure for political will is crucial, as is linking the environmental agenda with development priorities, especially for the Global South. The panelists concurred that integrating DPI language into environmental governance frameworks is essential for fostering a cohesive approach to digital transformation and environmental sustainability.

The event concluded with the recognition of the need for standardisation and interoperability in DPI to ensure its effectiveness and scalability. The discussion also acknowledged the importance of collaboration between the public and private sectors, as well as the role of open-source initiatives in promoting innovation and accessibility. The potential of DPI to significantly enhance environmental sustainability efforts by improving data sharing and transparency was a key takeaway from the event.

David Jensen:
Okay, I think we’re good to go. Welcome everybody. Welcome to this event on the future of digital public infrastructure and environmental sustainability. I will be your moderator today. My name is David Jensen. I’m from the UN Environment Program and I’m with the Digital Transformation Team. It’s a great pleasure to be here. But before I walk through our agenda, I just wanted to get a sense of who we have around the table because we’re dealing with these two kind of big issues, right? We’re dealing with digital public infrastructure on the one side and sustainability on the other. So I just want to know, hands up, who is sort of from the digital public infrastructure side that’s kind of interested in sustainability and sort of leaning into the sustainability side? Half the room. And what about the other way around? Who’s from the sustainability side that’s kind of leaning into DPI now? Fantastic. And we have anybody in the wrong room? Time to leave. Okay, great. So a good mix. And this is a very intimate setting. So we should probably have a really good conversation. So let me just walk you briefly through the agenda. We have kind of a three part agenda today. We’re going to open with a keynote presentation by our dear colleague, Rob, and he will provide sort of the key concepts and elements of what DPI is. So he’s going to try to demystify this for us and set the context. Then we’re going to move to two colleagues who are going to walk through a new UNEP report, which is on DPI and environmental sustainability. And then we’re going to go into a panel conversation with our four panelists. I will ask a couple of questions and I’m going to actually change the agenda. I’m going to rip up page two and we’re going to open it up for questions from the floor because it’s so intimate, cozy. I want to have a nice conversation. So page two is gone and we’re going to do a nice conversation from the floor. So with that, I think we’ve set the stage and I am very pleased to hand it over to Robert Opp, the Chief Digital Officer from UNDP.

Robert Opp:
Thank you, David. It’s a pleasure to be here. Thanks for the invitation. I have just a few slides. But maybe as that’s happening, oh, there they are. Okay, so one of these, okay, that one works, good. All right, perfect. All right, so as David said, I’m gonna sort of spend a little bit of time just talking about what we consider to be digital public infrastructure. It’s an evolving term, it’s a fast moving term. And then a little bit of the work that we do on using DPI in the climate slash environment space. And then hopefully that just tees us up for discussion for the report that UNEP has done, which is really a great contribution to this whole field. It’s, I’m only supposed to take five minutes. Five minutes, okay, well, we’ll try. Okay, first thing to say, what is digital public infrastructure? Okay, the most basic way that we try to explain this is in terms of physical infrastructure, governments can take the initiative to establish physical roads, bridges, airports, that kind of infrastructure. And the reason why they do that is because it’s what you build the economy on. So it’s available for the use of individuals, for the use of business, for the use of government itself. And so when we talk about digital public infrastructure, it’s basically that same thing, but in the digital space, what are the kinds of railroads and road networks and things that need to be established to really ensure that the digital economy is able to be powered? But when we talk about digital public infrastructure, there’s some key notions that really have kind of shaped this concept that come out of the G20 work last year under the Indian presidency. So they are these digital pieces of infrastructure, digital building blocks that are interoperable. that they are enabling access at a population scale, so they’re not just for a sort of niche, they’re available to all. They have to have governance around them and safeguards, I’m going to talk about that. And they invite also innovation and participation from different sectors. So the idea is that these are not stovepipes, these are platforms. And one more thing, because I have Yolanda sitting beside me. There’s an important concept here, which is as the reason why we support this notion of digital public infrastructure, is that we see countries that have pioneered this effort. So India, Estonia, Bangladesh, Brazil, others that have sort of found this combination and the sort of the way that this works. And countries have a choice. Countries can and have until now basically built this themselves, or they could purchase software technology from proprietary vendors. But there is a third path now, which we see the opportunity for these building blocks to be reused. And that’s openly available, open source, digital public goods can be deployed to build this. So for example, we start to see this exciting space of India, Estonia, and others that have, like, for example, Estonia’s xRoad platform has been implemented in many countries worldwide. And ITU, Estonia, Germany, Dial, are all, I thought I was going to miss one, are working together on this initiative called GovStack, which actually is also building new building blocks to be slotted into that. So I just thought it’s important to understand that notion as well. Okay. So why would we think about DPI when it comes to green transition or environmental sustainability? Well, every country needs to make progress on this. It’s a global challenge. There’s a possibility to scale quickly. this process, and we’ve definitely seen that in places like India where you can expand to hundreds of millions of people very quickly. And I’ll just give you a couple quick examples of where UNDP is exploring. One of them, when we talk about digital public infrastructure, some of the kind of core foundational elements are things like a national digital identity platform, for example, or national payments platform that really enables, facilitates payments between individuals, between companies, between others. If you’ve got that, those kind of pieces in place, then some of UNDP’s work that we call payment for environmental services, this is where we channel payments to farmers, for example, for conservation friendly practices. We want to incentivize them with the sort of cash in exchange for these kinds of actions and activities that they take. We can really facilitate that at scale very easily if you’ve got those fundamental layers in place. And so we’ve started to really build out these systems and look and experiment how they can work. And then the other area I wanted to mention is in the space of data and transparency. And this is where the UNEP report comes in with a lot of the kind of exploration of these concepts and how they can be deployed. You need to have transparently available data so that companies, government, others can know where the emissions are coming from, where the conservation activities are needed, where there are markets that can be explored. And so we’ve been working on how to support our partner countries in ensuring that they’ve got the data exchange or the interoperable data nationwide that can be offered and made available openly to people, businesses and others to take advantage of those opportunities and also to incentivize action. So we’ve been working on things like a transparency system that really makes those kinds of data available. Final thing I’ll say, and I know I’m probably over five minutes. I’ve mentioned it before, but a really fundamental component of digital public infrastructure is the right governance framework around it with the right protections and safeguards. That is basic form. I mean, think about any technology that you introduce without thinking through the issue of data privacy and protection. But there’s lots of different components to this. And we, together with the technology envoy of the UN office, are working on a digital public infrastructure safeguard initiative right now. And that’s working on all sorts of sort of different kind of building a framework of what that would look like universally. But there’s work to be done in the space of environmental sustainability as well, in terms of specific things that you would want to have in place to ensure that you’re doing sustainable procurement, that you’re using energy and renewable energy, and so on and so forth. And so it’s critical. And part and parcel of the DPI concept is that you introduce the technology, but you make sure that the governance is there at the same time. Otherwise, you really risk leaving people out and not having people in their rights at the center. With that, David, I don’t know how many minutes I went, but thanks for the opportunity.

David Jensen:
Well, thank you very much, Rob. I’m sure people will, at the question and answer period, take you up on that very last point about how do you deploy DPI and the governance framework in parallel. I think that’s a critical challenge. But let’s hold on that. Hold your questions for now. We’re going to go into section two, which is to go through the new UNEP report on DPI and environmental sustainability. So Anna and Kurt, I will give the floor to you. I am not going through the sort of typical biographical introductions because you can read the website and because we don’t have time. So I apologize for that. Thank you, David.

Ana F. Vergara:
Thanks. Well, let me start with saying that it is an absolute pleasure to be here. with you today to present our report, we are very proud. This is a report that is a result of a year and a half of research with contributions of over 40 international experts. As an interdisciplinary report, its main contribution, I would say, is bridging the conversation between different domains of knowledge, all on how to leverage digital public infrastructure for environmental sustainability, transforming our economic models. Now, everyone here probably has heard that data is a new gold and how it is critical to have accessible, timely, reliable, and insightful information that can support environmental sustainability and decision-making. But I want to focus today on one equally important point, and that is that in order to develop effective environmental strategies, we need to be able to connect three different types of information. Connect environmental data, economic data, and policy information. Allow me to further this point with a use case that we use in our report. So to tackle deforestation, we must look at what is driving those high levels of deforestation. For example, if we look at the beef production in certain countries or the consumption levels in others, that transformation and that information is critical to actually design policy interventions. We cannot stop there because we need to go further and utilize that same data to analyze if those measures are effective or not. Meanwhile, companies can also require environmental, economic, and policy information to substantiate voluntary and mandatory sustainability claims, especially when we’re talking about forest reports. Now, how to connect all of this data? Because they’re coming from different sources and there are different formats. They’re in the hands of different stakeholders. So this is why UNEP believes that DPI for Environmental Sustainability must be… must focus on enabling data exchanges. The monitoring of the environment, of economic operations and policies have all grown thanks to digital technologies. But despite this, we still find a landscape of siloed fragmented data systems, let’s say. And the more advanced examples of connecting all of this information can be found actually in private infrastructure, in private solutions. So for their own monitoring, basically, big companies have developed this infrastructure to gather relevant data. But what is the problem if we just rely on private infrastructure? We most probably won’t have access to all the sufficient information because they gather it under different business incentives, not necessarily for open sharing of that data. It may also increase data disparity because big players and small players will have different data. And it can also even increase disparity between regions of the world. Not to mention costs, of course. So a blend of digital, public and private infrastructure is needed. And what is UNEP’s proposal in this report? Basically, the report advocates for harnessing DPI for environmental sustainability as a way to enhance those data exchanges. That will improve data availability. And basically, this as a data exchange system will generate more data because it will create incentives and the mechanisms for that. It can also ease the discovery of the data that already exists and reduce the barriers for sharing that information. Now, we have many findings in the report that I would encourage you to read. But one of the key points is that a mind shift is required among decision makers to become active participants. under different roles, as data users, as data brokers, as data subjects, and one key point that I will finish today is that we need to make digital public infrastructure for environmental sustainability a priority. National digital transformation plans, but also on public investment programs. The world cannot afford to have partially useful information in the face of the climate, the biodiversity, and the pollution crisis. Let me stop there and I’ll hand over now to Professor Stockinger.

Kurt Stockinger:
Thanks very much, Annette. Hello, my name is Kurt Stockinger. I’m from Zurich University of Applied Sciences, and I will basically tell you now what kind of technology innovations we need here. I think I’m also allowed to share, so I will do this. So, basically, we’ll go through six different types of technology innovations, and I can only, since I only have four minutes, I can only give you some key ideas and some solutions, and the rest you have to read up, but just to show you a little bit some details what we need here. So, let’s start with open data discovery, and we all know the Internet is huge, has a lot of information, but also a lot of disinformation. So, now the big question is how do we actually find relevant information, you know, information that is relevant for your organization, or for your country, or for your use case. I think, now, are you sharing now? No. We see the DPI diagram with the six, it looks, we’re following the narrative very well. Okay, I tried to share again. Do you see this slide now? No. Open data discovery. You need to put the Zoom on the screen. Actually, can you just stop sharing, Kurt, and then… Okay, I’ll stop sharing. Okay. Okay, good. So, this is technology. I’m 10,000 kilometers to the west of you. I’m in Seattle, but we know how to handle it. So, basically, if you go to the next slide, please. So, they opened the data discovery slide. So, basically, the idea is how do we find the information and some of… Are you on the next slide already? We’re experiencing technical difficulties, Kurt. Just give us a second. Okay. So, maybe then I’ll tell you what is on the slide that you see next. So, how do we actually find the data? And one of the solutions is that we say we build a federated open data system. So, federated means that it’s not just one central organization or one central government or one central company, but it’s split into different parts. Everyone can own their own data and we can share it together. So, now the question is, how do we share it? How do we integrate it? And here we need some kind of technologies or what is also called the knowledge graph that you have maybe heard of, or basically a framework to integrate this. If we then go to the next slide. Technology work. Can you still hear me? We can still hear you. Give us 10 more seconds. We’re trying to sort out the technology. Okay. So, then maybe while you sort it out, I assume I go to the next one. So, once you have shared the data, the next thing is, how do you share it? Let’s say in a privacy enhancing way. So, because you don’t want to share all the information, you don’t want to know everyone, all the people in the world want to know what you’re currently eating, where you’re living, what you’re doing. And the companies are the same. They don’t want to give out all the information. They want to know what you’re doing. The companies are the same. They don’t want to give out all the information about the supply chains, for instance. So, you want to give out something, but not everything. And here’s some technologies that you can use. You can use encryption. So, you encrypt some of the data that is really critical and the rest you leave. Or you can use what is called some differential privacy. Yeah, we’re almost there. Two more seconds. slides and we’re on the slide that I want to. So this is about sharing. Yeah, this is good. I discussed this. Next one. I have some beautiful picture about encryption. So now we go. Thank you. Technology is difficult to handle. So that’s why here you see this nice log. And by the way, all these pictures are generated automatically by generative AI. So this is OpenAI or Dolly, whatever. So basically, you want to make sure that the data that you want to give out is encrypted. Or if you want to have some of it you want to give out, you can what is called differential privacy. So you add a bit of random noise to the data that it keeps the major distribution, but you’re not able to identify people or companies. If you then go to the next slide, now we have stored the data, we have encrypted it, we make it private. The next thing is, how do we make incentives for companies and for people actually to share their data? So we need a data market. And the data market should be fair. So it should not be that the big companies, they make all the profit. And the small people, they just give the data and they pay with it. So we really want to define here a data market. Like you go to Amazon or you go to eBay and you offer something on there. You want to sell your chair and people can buy it. So we want to do something similar here. So basically, how to make incentives for companies to share the data, even monetize, or even sometimes you don’t even need money for sharing it, but you give some credit points. This company shares the data, it gets a label or whatever. Or you can use blockchain technology where you can share the data in a secure way and you do fair transactions. If you go to the next slide, then now the thing is, there’s a lot of laws out there about what is an environmental policy, what is the regulation in a specific country, what is the regulation in a specific organization, for instance. And you can have many people to read all the laws and at some point. they might be through there. And when you afraid everything, then the law is already out of date. So this doesn’t really scale. So we want computational law. So basically, the whole idea is, how can we make laws human readable? And it sounds a little bit like science fiction, but it isn’t. For instance, what we are currently doing in databases, we’re doing something similar. So we store information about people. We store information about companies, which people work in which companies, or which products are produced by a specific company. You can do something similar also with a law. Basically, you can encode the main things in law in something similar what is a data model. And in a data model, you can describe relationships, for instance. So this is one thing. If you go to the next slide, how do we do it? So one way of using is large language models. And we have heard they are used everywhere now. You can draw beautiful pictures. You can write poetry. And you do many artistic, nice things. But we don’t want them to hallucinate. So we want these models, and especially if we talk about laws, we want them to say the truth, and not something which is made up, which sounds beautiful, but isn’t. So therefore, it’s important when we use these large language models or this AI technology that we always have a human in the loop. And if you go to the next slide, basically, this means we don’t completely rely on the AI to tell us what to do and to make our decisions. But we see them as a co-pilot. They provide recommendations. And we, as humans, in the end, we still take the decisions. I give an example. Every one of you knows how to use a calculator or maybe a computer. You wouldn’t use a calculator for everything. But if you have something to calculate, use the calculator. And you can still verify whether it’s correct. And here, we want to do something similar. So don’t rely only on technology. Still use the human brain and find a way of interacting. Now, we go to the next slide. So now, the big success factors of how we can do it, we heard before already that we need a transparent design. And we don’t want only companies to do it. But basically, we need open source. technology, because if we build this technology, we want to look into it and we want to understand it. It shouldn’t be somewhere in a secret place of a company that we don’t really know, so that’s why we want open source technology. And what is the major importance? If we want to have this endeavor, we need several people working together. Governments, they are great and they have great ideas, but sometimes they talk a language that data scientists don’t understand or they don’t even bother. They might say, oh, that’s politics, I don’t care, I’m a scientist, I want to do the real hard facts. So governments need to learn a little bit the language of the scientists so that they talk to each other properly. On the other hand, data scientists, computer scientists, domain experts, they also need to understand the language of the policymakers. So it’s really, it’s an important collaboration, because if we don’t collaborate, nothing really will work. We have some great ideas, great big words, but if the scientists don’t understand it in the other way around, we will never build anything. So that’s why it’s important. We need to tackle this together. It’s a very interdisciplinary challenge. Read the report, maybe hand it out, and let’s tackle the things together. I call this the six grand challenges or the 10 grand challenges, and we need to solve them together. This is open up the discussion now, and thanks for listening. And yeah, here’s the QR code for the report.

David Jensen:
As you say, the QR code for the report is on the screen right now. So if you want to have a look and download it, please go ahead. Thank you very much Kurt and Anna for wrapping up the second phase of this conversation. We’re going to move now to the panel, and we’re going to have our four panelists discuss some of the points that were raised in the report and some of the big challenges that you see in the field in terms of bringing together sustainability and DPI. So maybe, as I said, I’m going to start with the first round of questions, and then we’re going to open it up for a round of audience questions, if that’s OK with folks. So I’m going to start. with Sally Radwan, UNEP’s Chief Digital Officer. Sally, I think Anna mentioned the idea that we need data markets and we need data standards. Can you tell us a little bit about this big vision that UNEP has for the global environmental data strategy and how, if we actually materialize that vision, how that’ll reinforce and kind of support the work that you’re releasing today?

Golestan (Sally) Radwan:
Thank you, David, and good afternoon, everyone. Thank you for being here, and apologies for my voice. Can you hear me? So, as Anna said, and as the report outlines, and I really like the way you phrased it, we cannot afford to have partially useful data to tackle the big challenges. And this is probably most true in the environmental space. I don’t know how many of you have looked at environmental data. And at UNEP, when we say environment, by the way, we talk about what we call the triple planetary crisis. So it’s not only climate, we talk about biodiversity, deforestation, chemicals, pollution, et cetera, et cetera. And we’ve done an exercise over the past year or so to try and understand the landscape of standards. We also tried to understand the data landscape, but also the standards landscape. And we found more than 130 different standards for environmental data. So if you imagine that number, if you imagine a topic as complex as DPI, where we really need to link, as Anna also said, the environmental agenda to the development agenda, to the economic agenda, and you have 130 standards to deal within the environmental space alone, that becomes an impossible task. So what we’re doing at UNEP, also as part of our mandate from our General Assembly, is what’s called the Global Environmental Data Strategy, which is an attempt to explore and harmonize the environmental data space. And so- so the horizon is until the end of next year. So, so far we’ve converged on five pillars that we’ve agreed with our member states on what we really want to deep dive on and see how we can exchange best practices, but also what new initiatives we need in those five areas. And I’ll mention them briefly, and then we can elaborate in the Q&A if needed. So the first, of course, as Quote also mentioned, data governance. So how do we collect, manage, curate, update, secure, classify environmental data, which is a huge impediment, by the way, to some member states being willing to share their environmental data because there’s no classification mechanism there that allows them to distinguish between what’s shareable and what’s not. The second is particularly zeroing in on the issue of quality. And there are multiple facets to that, not only that there are varying levels of quality across environmental data sets, but also that we don’t really have an authority that can give a stamp of approval to say, this is a data set that you can use for such and such application. So again, exploring that space and seeing what we can do there. The third one is a big one, interoperability. As I said, more than 130 standards just for environmental data. So what can we do there? Do we need to converge around maybe five or six standards that we adopt as the main standards that we use for things like API, for example, when it comes to the environment? And then a really important one, especially for the Global South, is access and affordability. And that includes, of course, licensing issues, financial issues, open data, as Rob was saying earlier, but also things like machine readability and AI readiness of these data sets, which. wasn’t an issue five years ago, but now it is because we have to get ready for the AI age that started already a few years ago. And then finally, of course, capacity building. So how can we develop a comprehensive capacity building framework to help countries identify the skill gaps they have? Again, as Kurt alluded to, it’s not just the data scientists, it’s the environmental, the domain experts, it’s the judiciary, it’s the policymakers, it’s even the students and the general public. And they each need a set of skills and knowledge to allow them to play a role in this complex ecosystem. So this is roughly what the data strategy is about and happy to elaborate later.

David Jensen:
Perfect. Thank you so much. And Rob, as you said, that’s kind of a core sort of foundational building block, the data standards. And this is a nice segue now to Yolanda Martinez, the head of the GovStack initiative with ITU, because you’re now sort of moving up to that middle layer. And maybe you could reflect a little bit about what kind of challenges are you seeing in kind of bringing that sustainability dimension into GovStack and maybe reflecting a little bit of what you heard from either the presentation or from Sally’s presentation.

Yolanda Martinez:
Yes, definitely. First of all, congratulations. I thoroughly agree that it’s not easy to put together the expertise of so many people contributing. And I think this report comes very in hand to the global DPI agenda and ecosystem. And recapping of everything that has been said so far, I would like to put GovStack as an implementation layer. We provide what is needed to make DPI an enabler for many digital government services sector agnostic. At the level of implementing, for example, digital service to facilitate compliance with an e-waste regulation, which is a very… common policy implementation on a circular economy basis. You need a digital service that identify electronic producers. You need to set certain requirements to grant them certain fees to have a circular economy policy in place. That’s a digital service to facilitate compliance with e-waste policy. To enable that digital service, you need building blocks. To put together those building blocks, governments have different options. They can go open source for their digital identity software. They can do a tender. And for that, you need a technical specification. At the very granular level, you need a technical specification that describes key functionalities for how my digital identity is going to be managed. You need API’s definitions to connect digital identity with a payment gateway if you need to collect a payment, example that also Robert used. You need a workflow engine to define the rules for that data service to comply to certain regulation. So GovStack has 11 technical specifications for the most common reusable software components of any digital service. Identity, payments, scheduling, messaging. And the objective is to make it very easy for any digital team to digitize sector agnostic digital services. Together with that, we have a sandbox, which we aim to collaborate with different governments and UN entities to have sector-specific use cases. Leveraging our technology agnostic technical specification, then you need to have the human eye, as was presented, with experts from different sectors saying, okay, for our environmental ecosystem, we need services that facilitate compliance. We need services that facilitate multi-stakeholder participation. We need digital services that enable AI usage to better leverage data management and analytics. So that’s when sector-specific use cases can be. to leverage the technology that is available in Los Angeles. And the third element is capacity building, which I think also resonates a lot to what has been shared so far. We’re gonna launch tomorrow at 4 p.m. I would like to welcome everyone to the launch of the GovStack e-learning hub. The objective there is to have different e-learning materials, but also tutor-led courses like the GovStack Artifacts to leverage many expertise across many different sectors to become aware about all our offering or toolbox and to work with different use cases. We launch Women in GovStack Challenge and it’s the beauty to see how technology agnostic tooling can really enable across a variety of use cases from environmental, from health, et cetera. And finally, sector specific, we have two e-learning courses, Green Data Center Guide that we produced together so GIC and the World Bank and the Green ICT Procurement Guidelines. And we also have two e-learning courses related to that. So first of all, thank you, because I think there is a lot of resources already available and the beauty of having this type of knowledge products like that report really provides a sector view of the relevance of TPI. And I think GovStack brings what makes it very tangible and easy to implement.

David Jensen:
Perfect, thank you so much, Yolanda. Now we’ll move over to Meena. Now, I think Rob and Anna both talked about the idea that this is gonna need to be a blend between public and private sector. And whenever there’s a blend between public and private sector, there’s always gonna be a need for safeguards. So maybe we could pitch the question of safeguards over to you and discuss, what are the safeguards we’re looking at and what has been your experience so far in looking at some of these safeguards of TPI and sustainability?

Menna El-Assady:
Thanks for having me. I’m really glad to talk about these topics. So I’m coming from a research perspective. I’m a professor of computer science at ETH Zurich. We work a lot on explainability, bias detection and AI, but also human-AI collaboration and how to facilitate that and in terms of different types of interaction modalities. So. I think we already heard a bit about safeguards in terms of data privacy, but also like hallucinations of language models, for example, and how we can actually interact with AI and so on. But I think there are multiple layers of safeguards from knowing the value of data. There’s a lot of different concepts around how to actually define data, how to define data value for individuals versus organizations versus the public and so on. In the medical sector, people have been pioneering this notion of data donations, for example, so people know the value of the data that they’re contributing or even monetization based on what they’re sharing. The other aspects are figuring out representation in data. Right now, we are seeing a representation crisis, if I can call it so, in terms of the models that we see being trained and used for multiple different problem solving issues, being very biased, very centered around specific demographics or specific countries or specific use cases. I think we need to widen that in order to actually have representative data. Part of the work that we do is on bias detection and AI, for example, and there we can see, regardless of how you would define a demographic, a lot of different biases, and it is a very tedious, complex problem. When you say, I will de-bias a model based on gender, say, but then you add in a specific other factor that you split the demographic with, combining two of these factors already creates a new bias issue. When you think about the data that we’re collecting, the standardization, all of these issues and all of the different perspectives that we want to take on these data, since then, this becomes an issue. I think just to motivate a little bit what we can do, these are problems, these are things that we are researching, but just to give it a bit of a positive spin, as a university, we are doing three different things. We are doing teaching, we’re doing research, and we’re also doing development and interaction with the public and private sectors. So in teaching, we have courses every semester on 200 students, and they have to create full stack prototypes. So they develop complete applications. And we’re partnering with a lot of different organizations, partnered with like the UN OCC, the digital libraries at the UN, and a lot of different public and private sector organizations. And they pitched an idea where they need human AI collaboration, or like full, fully fledged applications. Students work on that for an entire semester, and then they we get to deliver different like proof of concept applications. And then students get to develop that further. So this is something that has been going on already for three years, there’s a lot of success stories coming out of that. The second is doing research for a very specific topic. So we can actually collaborate on a research topic to try to investigate how, for example, to kind of figure out standardization, how do we do data annotation? How do we bring humans in the loop? How do we have a safe understanding of AI and usage of AI and so on. And the last part is basically enabling platform development. And this is an initiative that we’ve started based on 10 plus years of experience doing these types of platforms. Right now, we’re in the process of developing an open source standard platform, where all of the researchers working on the topic of human collaboration would be able to contribute their particular modules and code to so that others can use that for rapid prototyping. So say you have a particular idea that you would want to do, you can kind of use this platform, and we’re trying to develop a lot of different kind of interface for it. So you don’t need to even code yourself and develop your own human AI platform and code. So there are lots of different opportunities for the future. We’re currently in the process of developing this further and also developing a web presence for it so that people can access it further. But I’m happy to talk to everyone here around the potential of collaborating on these topics.

David Jensen:
Perfect. Thank you so much. And I’m sure you’ll get to a lot of questions afterwards, even after the session. Final question, and then we’ll open it up for further questions. I want to to bring in Noemie from BMZ. And I wanted to connect the dots to what you said, Rob. One of the benefits you said was scalability, right? So once we get the standards in place and the applications in place, how do we scale this up? So what kind of approaches do you think we can look to to scale up some of these best practices? And can you reflect on what BMZ is doing to take some of this into its programs?

Noemie Burkl:
Yes, well, thank you very much for having me here. It’s really fascinating to see what experts can bring together and from the different perspectives. I wanted to talk about GovStack when you talk about scalability, because it really is our response. But since we have our GovStack expert here with Yolanda and also Rob, who already talked about GovStack, I’m not going to duplicate that as well. But I do think because of the interoperability, sorry, I always struggle with that one, it really has the potential of scaling up. And it is one of the best approaches that you can do. From a government perspective, I’m not an expert, so you will not hear the real technical words from me. We see that DPI has been tremendous to really contribute to this planetary triple crisis that you’ve just supported. And we are really happy to be able, and through GovStack, to contribute to having one complete approach to that. So through these digital building blocks, we really think that this can be transferred from one context to another on a country level. And this is something that we’re quite proud of as well to be also part of. So maybe one use case to make it concrete and not to take too much of your time, of these building blocks that I could mention to make it more concrete, where we have this applied, is the tree tracker. I will just briefly explain how this works. The tree tracker software simplifies tree management. It streamlines the process of distributing seedlings. which encourages more tree planting activities. Just to make it a bit more concrete in this discussion and to seize what we mean. This helps in an increase in green cover, which is vital for capturing carbon dioxide and mitigating climate change. When we are able to accurately register and track these different tree species, Softrek can ensure that a diverse range of trees are planted, which of course enhances biodiversity and increases resilience. This is just one very concrete example that I wanted to bring in to show what we mean when we talk about the concept of DPI and how it could link up to climate. And one additional thing that I wanted to mention, because also Kurt was talking, I mentioned collaboration and standardisation. He also mentioned open source, but that’s not what I wanted to talk about. Collaboration is really key and this is why I think it is crucial that we’re sitting in together with UN agencies and also implementing agencies such as the GIZ. So what we’re also contributing to is the DPI and climate knowledge exchange series that I wanted to mention. And that’s the GIZ, that’s our implementing agency that is participating in it. And it is part of the CODES impact initiative DPI for planet. That Robert knows as well, very well. It’s co-led by UNDP, co-developed the Rockefeller Foundation, USAID and GIZ. So they’re also, they explore how DPI can then be adapted and scaled to meet climate needs. So these are just two things that I would just throw into the discussion to have one very concrete example maybe, and also to underline the collaboration efforts. Thank you.

David Jensen:
Thank you so much. And that’s a really nice way to wrap up the panel. Now, we are at close to time, but I do want to have some discussion from the floor. So I’m going to take chair’s prerogative and extend the session by 10 minutes if that’s okay with people. Because I think it’s important that we have an interaction, interactive discussion. So we’ve heard from our panelists, we’ve heard from the report, we’ve had a scene setting. What kind of questions do you want to fire at any of our colleagues here at the table or online? Hands up if you have a question. Sir, please introduce yourself.

Audience:
My name is Harm Aronsdorst, I’m a Dutch from the Netherlands and I have my own company and I’m trying to have or support the trust on the digital society and economy for the last 25 years. And we recently have been working on a interoperability testbed together with the European Commission for identity wallets. So one of the foundations that has been referred to in DPI to have a trustworthy identity layer and then building up the cost stack or the DPI architecture building blocks. And what we found is that it is really helpful in creating a more ecosystem based building blocks that is always available. So interoperability testing of wallets, working with government services is a good thing to have. There are different standards in each segment and domain. So we talked about sustainability, how you monitor or measure or certify that from a lifecycle perspective or an environment perspective or a building or product perspective. But then when you go to the trusted data layer, I think there’s a lot of standardization still that is a mystic area. So how we can help and contribute to create a more safe and seamless, like a sandbox environment to experiment what works, what doesn’t. doesn’t, and how can we test and verify together is maybe a good example.

David Jensen:
If I can just paraphrase or summarize, maybe that’s for you, Sally. How do we engage the private sector in this global environmental data strategy and how do we begin to understand private sector needs both to contribute data but also to use data? And I think the question also obviously goes to Jolanda as well about sort of GovSec, but you want to take this question of private sector engagement and consultation. I’ll be very quick so Jolanda can also check in.

Golestan (Sally) Radwan:
So from our perspective, we’re obviously conducting consultations also with the private sector, but I think one thing that has also been flagged by our member states as very valuable is best practices. Just by being able to consolidate what’s out there and what has been done to tackle these different issues and give ideas essentially to governments but also to other private sector companies, I think having that basis of knowledge exchange in something like the report or likely different reports that the strategy is going to produce is probably a good first step. And then based on that, we need to develop certain frameworks to be able to test together as you mentioned. And just to clarify, Ana, you said there were 30 or 40 experts contributing to the report. Private sector was involved, right? Yes, specifically some institutions that were working with C-level executives so that we got their participation and their consultation to integrate how they view the connection between all of these different domains of data and how that was relevant for data public infrastructure.

David Jensen:
Thanks. Yolanda?

Yolanda Martinez:
And also to complement all the work that we’re doing, all the technical specifications were drafted with technical working groups and members of those technical working groups come from an open call process that we do and we have experts. members from public and private sector, academia as well. We launched during the DPI high-level segment on Monday the Open Wallet Forum, which is scaling what the Open Wallet Consortium was doing, but really leveraging the platform the ITUS standardization process can offer, which is completely multi-stakeholder and open for not only member states, but completely a multi-stakeholder process to participate in drafting the technical specifications. So very happy to continue that conversation with everyone here to actively participate.

Audience:
I’ve been supporting from the launch of the Open Wallet Foundation in Davos two years ago till today. And we see the momentum now also with the UN and ITU to have the Open Wallet Forum launched. I would just refer also to the initiatives, mainly by MOSIP as a COFSTEC implementation maybe where there might be relevance in just seeing how collaboration can be further reinforced.

David Jensen:
Thank you. Let’s move on to any other questions, please introduce yourself and if you can, if there’s a specific panelist that you’d like to ask the question to specifically, please go ahead.

Audience:
Hello everyone. My name is Fatma. I am an international relations student at Webster University. I also have the honor and privilege of working with Yolanda on COFSTEC. I just have a… Is this a planted question or? It feels like a setup. It’s a question that Yolanda has no idea about. I’ve been working with Yolanda for about two, three months on COFSTEC and one of the things that I’ve noticed within COFSTEC is that… there actually is political will to digitize, but I also know on the environment side of things, one of the big challenges is political will for countries to actually take action on climate change. So I was just wondering, how is this sort of tackled this issue of political will and does incorporating DPI make it easier to get governments to take climate action, especially because there is, at least from what I understand, create a political will to digitize than with climate action?

David Jensen:
Well, that’s the question of the year. Fantastic, thank you, Fatima. That’s a magnificent question. Who wants to take that question? It’s a big one.

Robert Opp:
I mean, I have one in my mouth.

Audience:
Cheers. Do you want to start it? Sorry, can I add something on the same?

David Jensen:
Perhaps you can add that question. So what we’ll do, we’ll do this. You can add to it. And then I will give each panelist the final two minutes to make a comment. And then we’re going to wrap up because I can see people are poking their heads in and they’re wanting us to sort of vacate. So add your addition and then we’ll go down the line.

Audience:
Yeah, I’m called Benny from Tanzania. I’m working for the government of Tanzania. So my question is straight for the GovStack. What is the strategy to speed up the implementation so that we can see the impact of this initiative to the general public? Perfect, thank you.

David Jensen:
Okay, so we have two questions on the floor. Rob, let’s start with you.

Robert Opp:
Well, I’ll try to address Fatima’s question very quickly, which is not specific to the climate action sector, but more in general terms. I think it’s fair to say that when we see governments starting to create the ecosystem approach, you very quickly start to see the expansion of that ecosystem. system approach. In other words, it makes it makes it so much easier to build on new layers of digital public infrastructure, because you’ve got the first ones established, that you start to see this explosion, especially if you open it to the innovation ecosystem. So, you know, you create your digital ID platform, and you, David, there’s a sign, a sign that seriously appear in five minutes. So I think, you know, what it becomes, like, you definitely pick up momentum as you go along, if you can get the fundamentals, right. So I hope this climate sector is the same.

Yolanda Martinez:
And I will complement to that, the importance of data driven policymaking. So I really celebrate the report that has been launched today. And I think it’s on us to first make aware all multistakeholder community that these knowledge products exist, that all these initiatives about aligning standardization, collaboration with universities, joined initiatives like DPI, self-ware, and very concrete tooling to respond to Tanzania. The objective with CovStack is to really go from design to prototype in less than eight weeks. And we run this exercise during the challenge with 14 prototypes live, we’re going to have tomorrow three best projects presented by the participants of the challenge. So I invite you to join that session, everything we do is open source, everything can be reused as a whole purpose, including the learning objects, you can incorporate the learning courses into your own digital academies, if you wish to, we’re really pushing forward to democratize access to all these knowledge products, very tangible resources and partnerships to make DPI actionable in a very short period of time, never forgetting that this is about people, that this is about designing services for people, that this is about giving tooling to policy makers to better them.

David Jensen:
Thank you very much, Yolanda.

Menna El-Assady:
I think I just wanted to add to this point about data driven decision making, and want you all to imagine like a future where we can actually have an interconnected system where we can, like have data driven decision making support, both on a macro level, like for governments and like entities and so on. But also on the very, very micro level, say you’re going to the supermarket and want to figure out like buy these oranges or these oranges, based on like some sort of big question about like supply chains and like environmental impacts and so on. And you want real time feedback and guidance based on all of these issues in these micro decisions. And we’re currently working with the UN OCC and the peacekeeping missions to provide real time decision making support for peacekeeping missions in very specific areas. But I think as research develops in these areas, we can actually implement that on multiple different levels, if the infrastructure is available. And I’m very optimistic that if we can actually have standardization and build all of these technologies, then we can actually provide a much brighter future for everyone.

David Jensen:
Thank you, Sally.

Golestan (Sally) Radwan:
So politicians are human beings. Even though we may not believe it sometimes, but like any human beings, they respond to incentives. And so what we’ve got to do is create the right incentive structure for there to be political will, because political will doesn’t spring out of thin air. The reason there’s political will behind digitization is that they can see the business case. If they can see the business case in climate action, they will suddenly have political will. So and this is why I think initiatives like DPI can play a crucial role. But also, definitely, we always talk about it at UNEP linking the environmental agenda to the development agenda, especially for the Global South. It’s impossible and unrealistic to ask a government leader from the Global South to suddenly forget about all the priorities they have in terms of hunger and healthcare and all these kinds of things, and then suddenly focus on the environment for no tangible link between the two. So this is the big thing that we need to work on. It’s super difficult, but if we can achieve at least a percentage of it, I think that will go a long way in generating employment.

David Jensen:
Perfect, Sally. Final word.

Noemie Burkl:
A lot has been said, but yes, human beings, as you said, and human beings also means that sometimes those human beings that make decisions just don’t know. So if your question is, does which DPI help decision makers to take and generate political will, I would probably say, first we have to translate what DPI is. So this is one thing that I would say, knowledge sharing, translation of what we see and how we see it and how it’s linked. And then to add to that, and this is not to be underestimated really, I keep telling my team all the time. And the second thing also is to show how it works, win-win situations, etc. We talked about the data-driven policymaking, Rob, the Data for Policy project, right? So we have a very good also Data for Policy project that UNDP has done with some small support from us to showcase how that could work a tool. You can elaborate on that later if some people are interested. So that is a showcase, but also win-win in terms of, you mentioned the linkages between environment and development. I would even say, because we’re here with digitalization and the whole climate change discussion that we’re having, we call it twin transition, to show the revolutionary… aspect that we are having on climate at the moment and on digital AI whatever and to show that you have to think it together and that you can if you think it together really make a huge difference and leap and also try to address the dangers of potential risks. Thank you.

David Jensen:
Perfect and I just wanted to add to that because I think you were touching on this but I think these two communities are still somewhat divided and we need to start cross fertilizing across both of them so in the climate community in the climate cop or the biodiversity community we just start bringing DPI language into some of those conversations right now it’s fairly absent a lot of digital transformation language is still not really reflected in the environmental community and the environmental communities instruments of governance right so there’s a gap there and I’d say the same gap is on the other side and whether it’s DPI community or the National Digital Transformation Strategies or digital policy environment is often missing and so we need to work on both fronts to connect those communities together and to embed our various asks in each of those governing frameworks so that there’s so that integration that you mentioned is going to happen. On that note I thank you for your engagement I thank the panelists I thank you for the report. I thank us for extending us for about 15 minutes so thank you whoever’s waiting in the hallway.