Global Progress in the AI Era: How the global south can finance AI infrastructure on its own terms

Summary

Podcast Summary: This Week in Global Development

Episode: "Global Progress in the AI Era: How the Global South Can Finance AI Infrastructure on Its Own Terms"

Date: March 2, 2026
Host: Katherine Chaney (Devex)
Guests:

• Ala Muravat (Managing Partner, Sustainable Growth at 500 Global)
• Kate Callot (Founder & CEO, Amini, Nairobi)
• Jonathan Reed (Minister of Innovation, Industry, Science & Technology, Barbados)

---

Episode Overview

This special edition focuses on how countries in the Global South can finance, own, and benefit from their own AI infrastructure—resisting the pitfalls of digital extractivism familiar from past technological waves. The discussion explores financing architecture, equity, capacity-building, risks of dependency, and the meaning of digital sovereignty, featuring perspectives from entrepreneurship, venture capital, government, and tech infrastructure.

---

Key Discussion Points & Insights

1. Who Owns AI Infrastructure—and Why Ownership Matters

• Ala Muravat’s Takeaway from India AI Impact Summit (02:48):
  • Ownership is questioned everywhere; the current funding landscape is dominated by those seeking tangible ROI, not the long-term stewardship needed for public sector uses.
  • The real gap is not compute power, but institutional capacity: procurement frameworks, enforceable data governance, building government talent pipelines, etc.
  • “As with the Internet, core AI infrastructure risks being owned mostly by the wealthiest countries, unless regulatory and ownership frameworks are reimagined.”

2. AI as Core National Infrastructure

• Jonathan Reed's Perspective (Barbados) (05:19):
  • Barbados treats AI as a critical infrastructure akin to roads, schools, and healthcare.
  • Emphasis on ethical automation: not just productivity but also protecting jobs, repurposing skills, centering well-being.
  • "We try to find a way to leverage this tool and advance life at the same time... The same investment for people that we're going to put in place in AI." (07:40)

3. Preventing AI Extractivism in the Global South

• Kate Callot’s Framework (Amini) (09:12):
  • Extraction occurs on two main levels:
    1. At the value chain: Those working on data collection and labeling (often from the Global South) derive the least value compared to those higher up.
    2. At the infrastructure level: Data is exported, models are developed elsewhere, and then sold back to original data sources at a premium—analogous to the old resource extraction.
  • Not every country can or should be fully sovereign, but should have minimum viable compute—right-sized local data infrastructure (e.g., micro data centers in shipping containers in Barbados) to keep control of the most sensitive data and applications.
  • “Governments as first users create local demand and incentive, then developers can build local applications.” (12:15)

4. Financing Strategies—Blended Models & Local Capacity

• How to Move from Pilot to Sovereign Infrastructure (14:40, Kate Callot):

  • Blended finance is key: MDBs, philanthropy, sovereign wealth funds and government budgets should join to derisk and catalyze local investment.
  • Government buy-in follows demonstrable impact; early layers may need external (patient, risk-tolerant) capital.
  • “Once governments see well-being outcomes, confidence and further investment follow.”

• Three-Layer Financing Model (Ala Muravat) (17:57):

  1. Patient capital with governance teeth: Public or philanthropic funding structured so the government keeps some control and decision rights, starting with ownership and IP in term sheets.
  2. Institutional capacity funding: Invest in people and systems, not just tech.
  3. Maintenance financing: Budget for year 3, year 5, and beyond, not just launch.
  • Highlighted untapped resources: Islamic social finance ($3.5 trillion globally), diaspora remittances, and rethinking domestic public finance (“AI is not a separate budget line—it’s systems’ modernization.”)

5. Rethinking Development Finance Institutions (DFIs)

• DFIs Should Crowd-In Private & Patient Capital (23:30):
  • Examples:
    • IFC investments in foundational digital rails like M-Pesa.
    • African Development Bank’s AFIDA initiative channeling capital for AI/data infrastructure.
    • European Investment Bank incorporating ownership/capacity tranches.
  • “DFIs aren't just providing capital; their risk position lets them lock in long-term public benefit and sovereignty.”

---

Notable Quotes & Memorable Moments

• On the value chain of AI in the South:

  “You have a lot of workers in Kenya, the Philippines—extremely educated computer science graduates—working at the bottom of the value chain for the benefit of bigger companies.” – Kate Callot, (09:34)

• On reframing infrastructure:

  “For people, we’re going to put the same level of investment in AI as we have in health, schools, roads. It’s always ROI, ROI, ROI.” – Jonathan Reed, (07:34)

• On the patient capital need:

  “Patient capital with some governance teeth—ownership of data, ownership of IP built into the term sheet at day one.” – Ala Muravat, (18:00)

• On choosing partnerships and sovereignty:

  “Sovereignty... is a spectrum. For me, it’s ‘Am I in control? Do you share my values? Why are you choosing me?’” – Jonathan Reed, (26:02) “Sovereignty for countries isn’t about ‘ownership’. It’s about agency and choice.” – Kate Callot, (30:42)

• On academia’s role:

  “There's a huge call for academia here... Where is the intellectual leadership in an age of AI?” – Ala Muravat, (37:49)

• Test bed vs. testing ground:

  “Being a test bed for ethical automation is exciting—if you have the know-how. Otherwise, you risk being a testing ground, which is very dangerous.” – Katherine Chaney, (39:06)

• On nation-scale ambition:

  “If this is conceived from Barbados and shared with humanity, is this a thing we’d be proud of?” – Jonathan Reed, (42:00)

---

Key Timestamps for Major Segments

• 02:48 – Ownership of AI infrastructure (Ala Muravat, post-India Impact Summit)
• 05:19 – AI as imaginative national infrastructure (Jonathan Reed, Barbados case study)
• 09:12 – Extractivism and new resource models for AI (Kate Callot)
• 13:50 – “Minimum viable compute” visual and Barbados pilot (Kate Callot)
• 14:40 – Financing mechanisms and blended capital models (Kate Callot)
• 17:57 – Three-layer financing, untapped funding pools, and diaspora capital (Ala Muravat)
• 23:30 – DFIs, equity investment examples, and shifting how development capital is deployed (Ala Muravat)
• 26:02 – How governments negotiate commercial partnerships (Jonathan Reed)
• 30:42 – Defining/operationalizing digital sovereignty (Kate Callot + Ala Muravat)
• 35:01 – The pressing need for unbiased education, arbitration, and decision support for leaders (Jonathan Reed)
• 37:49 – Academia, research, and intellectual leadership in AI governance (Ala Muravat)
• 39:59 – Test-bed vs. testing ground, and a nation’s role in ethical proof-of-concept (Jonathan Reed)

---

Flow and Takeaways

This episode interweaves a practical, values-driven debate on building the foundations for sovereign, resilient AI in the Global South. There is consensus that the how—governance, capacity, and partnership agreements—matters as much as the what. Both the pitfalls (extraction, dependency, lack of local benefit) and solutions (blended finance, patient capital, public ownership, community representation, ethical guardrails) are laid out with real-world cases and blueprints.

By the end, listeners get a clear sense that financing isn’t just about raising capital: it’s about building the rules, trust, and skills to ensure that AI genuinely accelerates—rather than undermines—development on local terms.

---

For further conversation, next episodes will address the role of academia and intellectual leadership in AI for development.

Transcript

[Jonathan Reed]

Foreign.

[Katherine Chaney]

Defining forces of our time. It's reshaping how money flows, how services reach people, and who holds power. I'm Katherine Chaney, senior editor for special coverage at devex and this is Global Progress in the AI Era, a special edition podcast series exploring what happens when AI collides with the world's biggest challenges. In each episode we speak to leaders from philanthropy, government, civil society and the private sector who are responding to this technological turning point. They'll break down how AI could unlock breakthroughs in health, agriculture and education, and what it will take to avoid the risks of deepening inequality. We'll tackle difficult questions and spotlight promising ideas that will determine whether AI accelerates global progress or leaves more people behind. Artificial intelligence is quickly becoming core infrastructure, shaping how countries deliver services, allocate resources and build their economies. But behind every AI system is a financing decision. Who pays for it, who owns it and who captures the value. And for many low and middle income countries, the risk is clear that AI could become the next extractive industry. Today we're asking what would it take to finance AI in a way that builds. Builds resilience, ensures sovereignty and delivers real public value? Welcome to Global Progress in the AI Era. I'm Katherine Chaney, senior editor for special coverage at Devex and today I'm joined by three leaders tackling this question from very different vantage points. Ala Muravat is managing partner for sustainable growth at 500 Global, which is a venture capital firm that invests in founders building fast growing technology companies. Kate Callot is the founder and CEO of Amini, a Nairobi based company building localized AI infrastructure across the African continent and beyond. I love this line from a recent profile of her in the time 100. Callot's ambition is to help those across the global south participate in the digital economy not as data workers or call center operators, but as builders. And finally, Jonathan Reed. He's the Minister of Innovation, Industry, Science and Technology in Barbados and he's leading the country's push to digitize public services, expand digital ID and build the kind of national infrastructure that will shape how AI is deployed in practice. Thank you all so much for being here. So I'd like to start with you, Ala. You were recently at the India AI Impact Summit and you were part of a session that posed what might be the central question for this entire conversation. Who owns the AI infrastructure being deployed across the global majority of. So what was your takeaway?

[Ala Muravat]

I think well, first off, it's wonderful to be here and with two of my favorite people discussing, I think one of everyone's favorite issues. But I think the, what I walked away with is everyone is asking that same question and it seems like everybody wants to fund the model as long as they can kind of see the bottom line. Nobody seems to want to fund the public servant who's going to manage that model in year five or, and I, and I, I think Kate will be excellent to come in here. But I don't actually think the gap is compute. I think it's institutional capacity procurement frameworks, data governance that's actually enforceable, local talent pipelines that don't just train developers, but train people inside government who need to ask the right questions of those developers. And I think we all walk away with a question of like, okay, is it building locally versus globally or who owns it, who benefits from it? And if we follow what is the status quo? I think the answer is going to be the same as it was for the Internet, right? Regulatory really happened in two core spaces and it was really predominantly owned. That doesn't mean there isn't global benefit, but the most localized of that benefit was in high income countries. And at the same time you also want to push back against this idea that it's that binary, right? Because it does feel like a false binary. We can say, listen, it's actually not local versus global, there's ways in which we all benefit. But I just don't think the status quo sets it up for that. So the impact summit in India was wonderful because I think it was a global south country really coming and saying, we actually, we have the capacity, the interest, the political will, the regulatory determination to really be able to own, build and be a big part of this conversation. And I'm always excited when I see that happening because it's easy to say, okay, well we need to be doing more globally. And in global south countries, it's much better when a country stands up and says, hey, we're willing to do it.

[Katherine Chaney]

Absolutely. Speaking of a country standing up, I have long seen Barbados as a very interesting pioneer when it comes to leaning into this AI moment. Minister Reid, I wonder if you can tell us a little bit more about what you're doing. I know you're on the road right now, coming back from some important meetings where I'm sure you were discussing just that. But you're working to digitize public services, build out digital ID systems. When you think about AI, do you see it as infrastructure in the same way we think about roads or electricity? That's often how digital public infrastructure is described. And what does that mean in terms of how the government thinks about financing and procurement.

[Jonathan Reed]

Thank you so much, Katherine. And I mean, kudos to you pulling together this amazing grouping on a cool topic. Yeah. I actually have become thinking of it much more as infrastructure. And it's easy for me to do that when I pitch to the Ministry of Finance, where I make it seem more weighty than perhaps others see it as, but it's absolutely quickly becoming that. And I'll speak specifically about Barbados and then we could scale out outwards. I think Barbados is actually in a peculiar and advantageous position of being a small country, population of 300,000 people. And by its very construct of being a small island, it has a relationship with every individual in a way that you simply not going to have in a large country. Right. If you have a country of 25 million people or 250 million people, the intricacies of managing every single life is not as paramount as you feel it in a small island. So therefore, the disruption potentially of automation and AI is so much more distinct and in a way disruptive. But when you think of it from the positive view, it's so much more positive and so much more optimistic in that we have a structure where we've always treat human life and well being and labor as sacrosanct and we create frameworks where we protect labor, individual life and quality of life and quality of service and quality of job security. Right. So because you have a new tool, whether it be in microwave, a car or lnn, we're not going to displace your livelihood and displace your life as a result of using a new tool. We try to find a way to leverage this tool and advance life at the same time. Now it's easier for me to say that than actually then evolve it and actually use it, especially when the tool is still evolving the way that AI is very much evolving every single week. Right. With these new announcements being all the time, but it's in our DNA. So my relationship with the Cabinet and my conversation with the Prime Minister or the Minister of Labor is, look, I think we could actually be a global test bed for ethical automation. A global test bed for ethical AI. Not just in terms of how we use the data, but how does AI in the tool affect humanity? Because ultimately it's a tool. Right. And so I think once you do that, you start thinking about how do you scare people less and how do you help people more. Right. And I have to really, really kind of centralize. I was literally having this conversation this morning with the Prime Minister, regarding that. Right, so and so, okay, you want to be able to increase your productivity, we want to grow gdp, we want to be more prosperous. And we're going to use the latest and best tools to do that. Absolutely. But in pursuit of that, we don't want to displace 20, 30, 50,000 families that are now seen as no longer useful. Because I have this new tool. No, I have to figure out a way to repurpose whatever their best skill sets are for the purpose of being part of society. So the society grows with progress as opposed to just being disrupted and being shattered. Right. So this is us being very aggressive and all in on AI. But even more than all in on AI is really using our value set of who we are as people to then frame how we use this tool. So that's one. And so it as infrastructure is not at odds at all in my mindset with that kind of ethos. It's as infrastructure. It's the same way that we've invested very heavily in health or schools or road infrastructure. It's the same level of investment for people that we're going to put in place in AI. So it's, quite frankly, the easy economic pitch for me, I just pitch it always as roi, roi, roi.

[Katherine Chaney]

I want to go ahead and bring in Kate because I know Amini and I mentioned earlier, Amini is working across the African continent and beyond, Barbados being one of those places. But Kate, I want to return to something I said earlier in terms of this risk of this AI wave creating a new form of extractivism. And I know that's something that you are actively trying to avoid with your company. So can you talk about exactly how you're doing that? And maybe you can mention the work in Barbados as one example of that.

[Kate Callot]

Of course. Thank you so much for bringing me here with also two of my favorite people. And I think when we talk about extraction or the way we see extraction, we see it happening at multiple levels. The first one is across the value chain that's necessary to build AI system. So you go from collecting, labeling, transforming, cleaning up data, and then training models based on that data and then being able to sell those models. So that's the first value chain. When you look at this value chain specifically, a lot of the value today is extremely realized in the second part of the value chain, which goes beyond the data collection and beyond the data labeling. While the people who are working at the beginning of the value chain are usually the ones who are benefiting the least. By that, I mean, you have a Lot of workers that are right now in Kenya, in the Philippines and a lot of global sales countries who are extremely educated, computer science graduates, young people who are very enthusiastic about technology, but because of the lack of opportunities in their country in terms of being able to practice their skills, they actually end up working at the bottom of the value chains for the benefit of the bigger companies. So that's the first layer of extraction, the second layer of extraction, we've heard a lot the past couple of months, years, how really AI or data centers and compute are factories for intelligence. So basically what's happening is that you have data that's considered sort of like new natural resources, then compute or data centers are considered the factories, and then the output is intelligence. Intelligence in the context of agriculture, climate resilience, automation, digital public service delivery and such. But right now a lot of our countries, because of the lack of transformation tools in the context of compute, are still being left or being considered as a sort of natural resource, ground extraction of data transformation elsewhere. And then the models are sold back to your population at a premium. So we kind of balancing those two extractive risks and those two are threatening us to repeat a lot of the old colonialism and extraction models that are world as seen now when it comes to being able to avoid that. Truthfully, not every country can be sovereign, completely sovereign, meaning have a whole supply chain with regards to compute. That only depends on your country. But also not every country has the capacity to build their own data centers or compute factories. Not enough capacity to power it, because it requires a lot of energy, a lot of water, and not enough talent to be even able to use those tools. So when we start working with countries, we look at what is the minimum viable compute that a country needs to equip itself with. What does mean is it means that if you start with understanding where the data is, what format the data is in, then what do you consider as a country? Critical data. You would realize that the answer to the question is less in the gigawatt factories that a lot of the global north has been trying to push to us as the only narrative to build AI and more in the smaller megawatt, low kilowatt, which makes more sense because not every data actually needs to stay in country. Some of it can be sent out to the cloud, and the most critical part can stay in country. But then the interesting part we've seen happening is that as you start building this for government, as you deploy the infrastructure, you structure the data and you give government the first offtake of that infrastructure Meaning government now become the first user of the platform. They start seeing how really the value is being created and the value that unlocks. And then government now want to increase the investment so that they can also give that to the local ecosystem, which means now developers have a place to practice. Devcraft developers have the necessary amount of compute or access to some local tools and contextualize data to be able to build localized application which include the local context. So we already starting to see a shift where those developers who before didn't have the opportunity because they didn't have the tools locally, locally are actually now starting to benefit from these tools and building applications that are more relevant for their countries than the ones that are the most widely used in the world today.

[Katherine Chaney]

And just to help make this visible or help our audience envision what you're talking about when you talk about minimum viable compute. My understanding for example in Barbados is you're working on micro data centers in shipping containers. So that's an example of what this can look like. I thought that was a really helpful visual. I did want to ask you a follow up on the financing piece because you were just talking about how you can get to the place of government buy in, which is great and the goal, but in order to get there you probably need different players stepping up. I'd love to hear, you know, philanthropy's role, the role of investors with different levels of tolerance for risk or different priorities set on impact. How do you get to that point from a financing perspective of ultimately achieving government buy in?

[Kate Callot]

It has to be a mix. I think there are some governments who still don't see data infrastructure as critical infrastructure. Those some of the mindset is starting to change. But sometimes a government has the choice of investing into a road and a hospital. Data centers are not going to be the priority or data infrastructure is not going to be the priority. So how do you go about that? We have to find ways where we can use blended finance coming from the mdb, coming from philanthropies, coming from sovereign wealth funds, global ones who want to be able to support other countries and use this as an instrument to accelerate economic development in other countries. So now you're seeing data infrastructure becoming a mandate for a lot of these organizations which creates a mechanism that then de risks the investment that a government has to make from their own pocket into such infrastructure. And hopefully once the first layer is put in place, that's when governments can continue building out as they start seeing the value, as they start seeing how this actually has an impact on their ecosystem on their citizens, how this benefits their citizens and increase well being and focuses on the outcomes for their countries. They actually now able to build more confidence into the reason why they need to invest in that infrastructure and then continuing taking on the journey. But it's a big mindset shift that is now starting to happen. But I'm hoping that this year and the following years we're going to see more of these models coming in.

[Jonathan Reed]

Foreign

[Kate Warren]

hi, I'm Kate Warren, Executive Vice President and Executive Editor at devex. At devex, we don't just cover the biggest moments in global development. We create space to understand who and what are driving the headlines. Alongside gatherings like the World bank and imf Spring and Annual meetings, the World Health assembly, the UN General assembly and beyond, we host devx Impact House where our journalism comes off the page and onto the stage.

[Ala Muravat]

We bring together a curated group of

[Kate Warren]

leaders for live interviews, intimate roundtables, hands on workshops, and candid conversations you won't hear in the official meetings. It's where tough questions get asked, the spin gets stripped away and meaningful connections happen. If you'd like to join us or stay in the loop on all of our events online and in person, please visit devex.com events Ala I'd like to

[Katherine Chaney]

go to you because I know you have been very vocal about the fact that capital flows shape power. This was a big topic of conversation at the summit in India and that back to this point of the risk of AI becoming another extractive cycle. A lot of that has to do with the financing architecture. So can you tell me from your vantage point, how is the financing mix working to get to the kind of place where we can have this localized AI infrastructure that Kate is talking about and that Barbados is trying to really champion?

[Ala Muravat]

Yeah, no, I think Kate touched on it excellently. I mean, I think the financing you need is actually kind of three layers. You need that patient capital with some governance teeth that Kate was mentioning. So think about capital that is structured so governments can hold some decision making power over the systems being built, ownership of data, ownership of IP built into the term sheet at day one. And oftentimes if you think about kind of a fund where you're leveraging some philanthropic capital for your risk capital and then you have your DFIs and MDBs coming and in your secondary tranche where they're interested in impact, but they also want to see that there's potential for returns and then private sector partners who are really excited about returns. But if you're able to create facilities like that, I Think it can be really, really compelling. Second, I would say you really need the institutional capacity financing. So that's the memory to train the developer. And not just the developers, but procurement officers, regulators, ministry staff who need to govern these systems. Right now, in most things I see that's treated as a side budget. And I really think it should be the first line item. Because a country with strong institutional capacity and mediocre AI tools will outperform a country with world class AI and no ability to govern it every time. And I think that's very important. And then I'd say the third is really maintenance financing. So a lot of people are interested in funding deployment and getting into new markets and getting new people online. I'd say less are interested in kind of what does it look like at year three, at year five, at year seven, when the system needs updating, talent needs retaining and training, infrastructure needs scaling. So if your financing model doesn't account for that, you're not really building kind of permanent infrastructure. I love when people compare it to kind of roads and hospitals because it really is the same. Like if you think about kind of the way in which you would want to fund that kind of infrastructure, it's there. And then the number one question I always get is okay, well where's the money coming from? Obviously those who have the resources will be the ones to put it into their own countries. And I actually think it's very interesting because the money is there. I'm not one of you know, I genuinely believe we actually have more than enough. I mean, development finance institutions right now are sitting on billions earmarked for digital transformation that's often being deployed as traditional grants or technical assistance. So not necessarily risk capital that can crowd in other partners. If you redirect that into instruments where governments co own the infrastructure, you've changed the entire incentive. And then you have sovereign wealth funds around the world that are actively looking and deploying to create, to support AI AI infrastructure plays. They have the patience, they have the skill. I think the opportunity is designing vehicles where that capital flows into public AI systems in frontier markets, not just into ones where they see capital returns coming sooner. And then of course, I would not be me if I did not mention, I think Islamic social Finance is this $3.5 trillion pool that most of the development community hasn't even engaged with. So you have Zikat, which is obligatory charitable giving, and we're in the middle of Ramadan. So it's very timely, but it mobilizes hundreds of billions of Annale have WUGF endowments that are perpetual by design, which solve a maintenance financing problem that every donor funded system eventually hits. And sukuk can be restructured around tangible AI infrastructure assets. Right? And these exist, these instruments aren't theoretical. They exist and are operational at scale on other priority infrastructure areas. And so if we think about AI in the same way, I think we can hopefully leverage that. Kate touched a little bit just on domestic public finance. No one ever wants to talk about that. But countries already spend on health systems, education systems, public administration, administration. I don't think AI should be a separate budget line. It's just a modernization of systems. Governments systems government is already financing. Right. And then of course the last one I would put in there is, I think there is a very interesting opportunity for diaspora capital and remittances. I think you have hundreds of millions flowing into frontier markets every day, largely informally. And I think if you could structure that in a way where it was tied to local AI infrastructure competitiveness, income uplift, job creation, I think that would be a really exciting capital source that's already deeply aligned with country level outcomes and with, with better standard of care and living for, for people in the country and for citizens in the country. So I think the money exists. I think the huge gap is architecture and I think we do have an opportunity to solve it. But sometimes we're kind of because of how quickly things happen in AI, everyone's looking at like you know, a month away rather than necessarily what, what do we need to have to make this realistic for three, five, eight years out?

[Katherine Chaney]

So as you said, the money exists, but the gap is architecture. And I wanted to build on one point you made specifically to see are there any examples that could point to the way forward? You mentioned development finance institutions sitting on billions earmarked for digital transformation. And I thought this was interesting. You mentioned that it's being deployed as traditional grants or technical assistance, not necessarily risk capital that can crowd in other partners. So, and it's I, I find it very interesting that, you know, you're with 500 Global, a VC firm, and we're talking about markets and development priorities that you don't often expect VC firms to be part of those conversations. So I'd love to touch on that point exactly how is the development community or development finance institutions missing the opportunity to crowd in actors like 500 Global with money to invest in these markets that could ultimately help drive toward development outcomes. But money's perhaps not being used strategically. And if that's the problem, are there any examples that Point to how we could solve that problem.

[Ala Muravat]

Yeah, no, and I need to be more generous with my words. I think there's so many in the development finance space that are actually really commit, you know, and in the institution development finance institutions that are really trying to do things differently. I would say you have, you have quite a few examples actually of where they've done equity investments. So things like IFC gave equity investments into African fintech and health tech platforms like M Pesa's infrastructure layer. So the investment wasn't just in the app, but in the Rails that other services could build on. If you apply that same logic to AI, you invest in the shared data infrastructure, the compute layer, the interoperability standards, and you've created a public good with a return profile that can benefit more people. Yeah. And then the African Development Bank, I mean they have an AFIDA initiative that's really exciting. It's been designed to channel capital into AI and data infrastructure across the continent. I think that ambition is actually very right. They have dedicated vehicles for foundational digital capacity, not just kind of like one off innovation grants. And then last one that we've spoken to quite recently has been European Investment bank which is really thinking about how they structure digital infrastructure deals to include local ownership provisions and capacity building tranches alongside capital deployment. So I think the pattern across all of these, even the ones that honestly all of them, the ones that work, the ones that don't, is that DFIs aren't just providing capital, they're using their position to set terms that protect long term public benefit. Right. And I think that is what can be really exciting. That's what a risk capital from a DFI should do, that private capital won't. It would use its concessional position to lock in sovereignty, not just de risk a commercial bet.

[Katherine Chaney]

Great. I want to make sure to ask a question about AI labs and their influence in this space and you know, specifically when it comes to financing and partnerships, how low and middle income countries can navigate really these new entrants, I think to a lot of global development work. So you see groups like Anthropic or OpenAI signing partnerships with governments, you know, with goals in health, education, digital public infrastructure. And you know, I think there's a mix of praise and excitement, but also some concerns about what we're talking about in terms of countries having sovereignty and agency and capacity. So I'd actually love to go to you on this, Minister Reid. When an opportunity or a deal like this comes across your desk to work with an external actor, what are Some of the key questions you need to answer before signing off on something like that.

[Jonathan Reed]

Yeah. So my, my phasing of this has evolved, if I'm being perfectly frank and we're friends, so I'll, I'll be frank. I think there is something that feels transformative around a highly developed, highly sophisticated telecommunications superpower that could come to your country and do something that is truly transformative. And there was a moment in time where that was very much what I was going after. And I was kind of the core of how I thought about the ability to kind of rapidly transform the state of development in this country. I have evolved a little bit. Well, a lot. And so I think of sovereignty differently to how I did previously. And I think there is nuance to the idea of sovereignty. And where I've landed at this is not a state position. This is, my position is, I think sovereignty is that you have to have a reasonable component of control and a control of ownership. I don't think it's going to be 100. I don't think. I think there's somewhere between 0 and 100 that you're going to land and for different people, you're going to feel a sense of comfort on it. I mean, Nvidia is the most, is the, is the most profitable country company on planet Earth. The founder of it owns about 2%, 3% of it, which has made him incredibly wealthy. Right. So he has control and interest and a lot of decision makers, but he doesn't own the majority of it. Right. So I think countries have to make decisions about their infrastructure. And when you talk about the range, where, range of places where finances can come from, you're going to have to split the ownership of the equity or the debts with others. And therefore, how do you split the ownership of the infrastructure and the data and everything that's related to it? I think individual countries are going to have to make that decision for themselves. So that's one. And so for me, it's really around, am I in control? Do I feel comfortable? Do you share my values? Why are you choosing me? Like I asked myself all these questions, like, you could be anywhere in the world. Why you choose? What is unique about me? What is unique about a relationship? How is it going to last beyond my role in this position or your role as my point person, who I might love in you, in the country and this company having this relationship? Are you going to share each other's values for the next 20 years? Is it built so that when I win, you also win? And when you win, I also win. How is it, how is it part of the ethos of everything? So that's how I think about it, right. So and so I spent a lot of time on it and that's honestly one of the huge advantages of being Barbados because it becomes really simple for me. I'm a small market, I'm not a large scale commodity market. I'm going to be a research high value place where you want to try something interested and new in. Okay, is the new interest and new interesting thing that you want to try, is it going to harm my population potentially? Is it going to break something? Is it going to disrupt if it's in some sort of ocean technology, is it going to pollute the oceans? If it's something regarding public data, is it going to be able to strip away the dignity of my population by having the data exposed on the dark web somewhere? What are the things that are risk? And if I mitigate the risk and say, hey, the upside is huge, then I can have a conversation about it. And then you have to then spend a lot of time on am I going to enjoy having dinner with you and hear your values and your values at hand, who my nation is and who are I as a leader so that I can trust you when times get tough? Because things will get tough, things will go wrong. Right? So these are the kind of things that I look for. Again, this is not a, this is not a government prescription. This is really my prescription as someone who has to make decisions. I have to present things to decision makers, what I go for. And so for me it's can we build architecture where I feel mutual share value and mutual share risk and if that's the case and all things being equal, then we can have a deal, right? And after that is then about deal4flow, right? Does it make sense? Is the timing right? Does the capital deck look good? What are you proposing? Does it make sense for me in this moment, in this pocket of time, et cetera, et cetera? But I really start at values and that's been very, very useful for me. Sorry to get a little bit more philosophical on a tech talk here, but that's how I think about it.

[Katherine Chaney]

So Jonathan, it's really interesting hearing this more nuanced view of data sovereignty, digital sovereignty. Your point that it's not an absolute, it's sort of a spectrum and there are lots of judgment calls and you to figure out where values align before you move forward. I'd love to hear from Ala or Kate, how you've navigated that same nuance in terms of digital sovereignty. And kind of back to the financing question, why the financing of these partnerships matter in terms of getting to a nuance that works for countries.

[Kate Callot]

So on my side, I think as a company, we define sovereignty very, very similarly to what Minister Re just expanded on. And for us sovereignty for countries is not about ownership, it's about agency and choice. So choice to be able to pick which data you want to stay sovereign or residents. Choice to be able to participate and make sure that, you know, you understand who controls what, who benefits from what, and to what level, and really who captures the long term value in the stack or the investments that governments are making. So when we think about this, we have have built a sort of framework and it's about allowing countries to understand how to access. So a lot of countries we're working with have not been able to access a lot of these tools before. So how do you ensure that, you know, you help them build the relationship, the partnerships. For them to be able to access the latest and greatest technology that exists doesn't mean they don't buy everything. It just means that at least they have the choice to. The second one is around participation. How do you make sure that, you know, those countries are not left behind and they're giving the opportunity to participate? Third one is connections. So looking at really how we make sure that there is equity and that the country benefits and retains some value, meaning economic value creation happens in country and not only outside. And then the last one is representation, representation in the data. The agency to be able to say, hey, if you want to operate in my country, you have to adhere to my values. You have to make sure that the technology you're bringing in is representative of my citizens. You have to make sure that the technology you're bringing in is representative of the challenges, the constraint, the opportunities and the beauty of our countries. So having that representation as one of the key component of the framework becomes important. And like you say, there will be spectrum around data sovereignty, but it's about really defining which data do I want to be able to control, which data do I want to stay sovereign, which data do I want to open? And it will look differently for every single country in the world. But giving them the choice in the agency, I think it's where we truly make sure that everybody can have a chance to participate. And for us, that's our entire mission.

[Katherine Chaney]

Ala, did you want to jump in on that in terms of your point on the financing architecture needing improvement, how does that play into this?

[Ala Muravat]

Well, I think so. I mean, right now a lot of these companies, OpenAI, Anthropic, Google Meta, they're all kind of moving into the development space, right? OpenAI is actively engaging directly with governments in Africa. Anthropic is positioning itself around safety and responsible deployment. And I think there is actually, in my opinion there's some real value there. These companies have built capabilities that no single country is going to replicate and no one is asking them to. But I think it gets interesting if a country is building its health system or education platform on top of these companies, on top of their API terms, the pricing model, the data handling policies, because even though we don't think they are, those are very big governance decisions, you know, whether the country realizes it or not. And a pricing change, a model depreciation, a shift in terms of service that's going to hit, you know, a health system, a national health system very differently. So the role I would want them to play is, is genuine infrastructure partners, long term pricing commitments, open model options, data sovereignty guarantees that Kate and Jonathan mentioned real investment in local capacity to fine tune and adapt models for the local context, not just, you know, many of the partnership announcements I see like API access. And I think the companies that figure out how to do this will be trusted long term infrastructure partners to governments and they won't be just vendors. And I think it will define this entire space. And the governments that negotiate those relationships with clear terms from the start will be the ones that benefit and set the tone for others. Right, so who's going to be the first mover on the government end and who's going to say to Jonathan's point, like hey, take it or leave it, these are our values. If you want to work with us instead of being enticed by kind of a shorter term partnership, I think is going to be very interesting to see.

[Jonathan Reed]

Can I just chime in a little bit on that? And I just wanted to connect both points I think. And it's not an excuse because Kate works a lot with me, Alan knows me well enough. I'm, I'm, I'm super, super impatient as a person. But then I think of the idea of what we're talking about is something that we would not been speaking about 10 years ago, right? So like this is very, very new for the average citizen, right? This is, the average citizen hasn't been working on deep tech research in AI for the last 20, 30, 40, 50 years like some very nuanced group of people have been and now become global experts on these things. And we're hearing them use the language that has become popularized on very complex terms that didn't exist before, right? So you take something around infrastructure, around health care, around roads, around airports, around ports, around housing, around education. These are things that existed for a very, very long period of time. There was decades of nuance, decades of experts being able to speak to the public in taking complex terms and making them general place. Now we're rushing ahead with things that, you know, hitherto didn't exist a year ago, now being seen as, you know, part of your furniture, right? So that's difficult, right? So you have people who have deep experience, deep knowledge, want to do amazing good for their country, now have to segment how they're thinking and a thing that they don't feel comfortable in, right? And the truth is some of them have a hard time saying, well, we don't know, right? We don't know who to trust because the people who are the experts actually selling us products too. So we don't know whether they're trying to sell it to products that is just beneficial to them or not beneficial to us. We don't know. So you have a situation where a couple of things happen, right? One where you make bad decisions because you don't know what you don't know and you make a bad decision or you make no decision because you don't know what you don't know, right? And so both things are detrimental. So I think there needs to be much more around the idea of education for leaders on AI decision making. I think that's a huge myth that doesn't exist, that there needs to be a kind of, of non selling party that allows people who are making really critical decisions around this thing to not being felt like they're, they're going to miss a boat if they don't make a decision or they make a bad decision if the person that they're buying from is just ultimately looking at the bottom line. I think there's a lot of that that's happening that we're not speaking towards and is, I'm not suggesting slow down because I hate the notion of slowing down on anything. But I think there needs to be a speeding up of education, is there need be to be a speeding up of how this thing affects me and the values of those things and so people could make better decisions so they don't feel like they're missing a boat or worse decisions that be made for them by others that then negatively affect them without them making their own decision. So I think, I think that's missing. That's certainly missing. And someone has, or some group has to be kind of like that arbiter of knowledge where I don't feel like I'm being sold, I'm being talked down to. And although. Although it's. Everything is kind of. How do I get my country forward using this. These new tools that are now resources.

[Ala Muravat]

I was just gonna say, I did not expect, you know, 30 minutes in for us to give such a huge call out to academia and research, but I think that's actually a really good call out for the role of academia in this. I mean, so much of what we're talking about, like everyone, to your point, Jonathan, everyone's trying to sell something. But if could you actually get some researchers, some academics who are, I mean, who. Who are looking at this from a theoretical perspective and its applications to say, okay, this is. This is actually what makes sense. This is what we've seen happen in other countries, et cetera. And I think it actually, you know, there's that book about how history always repeats, which I think is actually very fair because yes, AI is new, but I mean, before it was the Internet and before it, to your point, have been. Has been infrastructure that we've had across so many issues. Right. And so I do think there would be a really, really interesting, you know, I think very few conversations kind of pinpoint, okay, where is the intellectual leadership in an age of AI. And I think that's actually a very great call in for that.

[Jonathan Reed]

Them.

[Katherine Chaney]

Agreed. And we'll actually be having a podcast episode focused on exactly that. The role of academia.

[Ala Muravat]

Just pretend it was my idea. Exactly when you do it.

[Katherine Chaney]

Inspired.

[Ala Muravat]

Yeah.

[Katherine Chaney]

Some continuity in the, you know, I've

[Jonathan Reed]

spoken like a true vc. You know what I mean?

[Katherine Chaney]

Oh, gosh, it's really.

[Ala Muravat]

I, you know, the hats. The hat hasn't been on for that long, Jonathan, but it fits.

[Katherine Chaney]

No. Minister Reed, I, I wanted to bring up something you said earlier that I think think is. Is natural to revisit in this moment, because you mentioned earlier that Barbados could be a. A test bed for ethical automation and framing it kind of as an exciting opportunity. And I think it is an exciting opportunity when there is know how about how to navigate this space. But as you mentioned earlier, that know how is lacking in many governments. And then there's the risk of being a tough test testing ground, not in a good sense. And I think that's what we're talking about right here, essentially being taken advantage of. So I just kind of wanted to revisit that Again, I think for certain governments, being a test bed is exciting and for other governments, becoming a testing ground is very dangerous. Would all of you agree with that? Because I think that's an important distinction to make.

[Jonathan Reed]

Yeah, I like your language. So I'll give you a personal anecdote. I used to be Chief of Staff to the Prime Minister and my natural bent was to go searching for new things that add new value. And, you know, before it's kind of become commonplace now or very popular in the developmental circle that people know, our Prime Minister, Mia Motley, and the work that we've done on climate advocacy and climate finance and all those types of different things. But before that, no one knew Barbados, right? So I would be on a call talking about some interesting project I want to do and people are like, where are you? And I would find a way to weave Rihanna into a conversation about. About either it's space exploration or coral reef regeneration or something like that, Right. So this to give context. And so for me, it's like, we're not a large scale market, we're small market, small open society, market economy. And so the highest form of value is not the same value. If I'm thinking about the Philippines or Indonesia or Brazil or the United States or China or India, I don't have hundreds of millions of people or billions of people where you ultimately are selling your product. So my point of reference is really on the highest form of what I think is the highest form of value, which is closer to creation than it is for market sale, right? So for me to do that, I have to be at the stage where ideation is happening, right? Where the original thought is being conceived. And before the thought gets towards a commonplace market, it has to be tested, it has to be kicked, the tires have to be kicked. Someone has to decide to regulate it first. Someone has to make it available to the public, Public. And that's kind of where we put our hand and say, hey, we think we could be extraordinarily valuable in that space, right? So, but we don't. We can't be that for every and everything because every and everything doesn't suit our geography, our profile, the problems that we want to solve and how we go about things. So we have this barometer is, hey, does this solve a really important local problem or regional problem? Does it make sense to solve this problem from us? Does it align with the values that we have and things that we'd want to solve, solve anyway if we had a blank checkbook, right? Or if we had the opportunity to focus 100 PhDs in an area. Would this be the kind of topic that we would go after? Yes. Right. Okay, this makes sense. So solving non communicable diseases, using AI to solve problems around cancer or something around ocean warming or any of those things that are within the line of what we're thinking about anyway. Does that make sense for us? Yes, it does. Okay, so do we have enough experts that we have high level of trust and values for sets that connect with us to be the arbiter of the truth first? Because it's a topic that we don't know. Right. So are there points of trust we're going to trust the partner? Yes. But we also want a mitigation divide in the middle that is really set decided by us that if anything goes off kilter, that we have this body of truth, this body of persons that we can believe completely to say it, to stare left or right. And depending on how things are going to go, it's a topic that we have that. Yes. So if it has a topic that we have that, then we could proceed further. And then of course it's about deal flow, about finance, about the legal ramifications and the contract and all those things. And we stick to the letter of the law. But I think those fundamental things have to be in place for us to make a decision. Right. And then the last part I think about a lot, to be honest, is because we're small and ultimately the highest value of the new creation that we're going to have isn't ever going to be just for us. Can it be a thing that we'll be proud of to contribute to humanity? Right. If this is the thing that is conceived from Barbados and is then shared with humanity, is this a thing that we'd be really, really proud of? That this was conceived or worked on and developed from here and people involved in it came from here and they lived here two, three, five years and they mingled with locals and we had 100 or 200 people working in this new area. And there's a really story of that. We sense a sense of pride that we contributed to the planet. That's kind of like the last, last piece in the puzzle for us to really make great decisions around. And so we're all in on the tech because we think that the tech is not the problem, the engineer is not the problem. I agree. If Allah to a certain extent that the finance is becoming less of a problem. What is becoming more of a problem is like what are our values and staying focused on the tech not just for the sake of tech and the tech more for how do we move humanity forward. So you know I tease my friends by saying a lot less nano banana journal and a lot more AI for cancer right A lot less of actually how do we use tools to empower life to be better as opposed to just you know mucking around and making dancing cats.

[Katherine Chaney]

Well what better way to end than that? This is definitely one of those conversations I wish we could continue but I know you all are busy doing the work and have to get back to it but the choices being made right now are really going to determine AI's future for better or for worse and financing has a big role to play in who is steering that ship. So thank you Jonathan, Kate and Ala for your work in this space and for taking the time to join us. We really appreciate it.

[Ala Muravat]

Thanks so much.

[Kate Callot]

Thank you for having us.

[Jonathan Reed]

Thank you so much. Sat.