We faced this issue with there are well over a million Sudanese refugees in Chad and we had set up mobile clinics with the government of Chad to go into these refugee camps and do screening for TB because frankly whenever you get situations with large numbers of displaced people you end up with TB. And the question is not replacing any radiologists. There were no radiologists. And so if you want the screening to be interpreted, there is no alternative. So in some ways I think one of the reasons this may well take off faster in low middle income countries is because there won't be the resistance from people who say this has taken my job and I don't want to change the way we do things, because in fact it's compensating for the fact that those people don't exist. But I want to pick up on the horizontal point that Bill made which is one thing many medical systems in low middle income countries have excelled at is creating paperwork. There is an enormous amount in many healthcare systems of literally paper records with carbon copies and all this kind of stuff and that makes it very difficult to capture because it's on paper. It also just creates an enormous amount of time for overstretched healthcare workers.
Do you mean like records, patient records?
Oh, well, if you've ever gone to some village and the community health worker will proudly take out a big kind of thing like this and they've carefully written in the name of each patient and what they had and all this kind of stuff and then you think to yourself well a that took quite a long time. But b how is that captured in any kind of national disease surveillance system or so? And the big difference is that now in Rwanda, your community health workers have an app on their phone. It has the details of the patient. They put it in, it's all recorded. They can look at what happened last time. They don't have to go back through the files. And also from a national point of view, that data that's being inputted by the community health worker is being captured. So your community health worker, you've got like 60,000. They become your disease surveillance system, right? And that's incredibly powerful. And then once you've got that kind of level of data capture, you can be applying AI tools for pattern recognition of all sorts of ways of detecting perturbations. You know, this region is seeing something slightly different we didn't expect. So taking paper out of the system, capturing the data, I think that's incredibly powerful.
Can you do that? Can you guys at the Global Fund go and take all those paper stacks and make them?
We're certainly investing in it. I would say that different countries are at very different states in terms of where they are on that pathway. But we do face a challenge, right? There's less money than there was. I mean that is the reality. And a lot of the money that was being invested in things like disease surveillance has stopped.
Well, because major donors around the world have...
Well, it's actually quite broad-based. We've seen quite a lot of donors reducing funding quite significantly. And that means that the investment... and we face a challenge which is that... let me put this in perspective. We've actually just had our eighth replenishment. We've just raised just under 12 billion with some donors still to go. Actually that shows remarkable commitment in this context by the donors who support the Global Fund. But we are going to be allocating less money to countries and they have had sharper reductions from other sources. So countries are going to face some very difficult choices. Do they invest in underlying system things like disease surveillance and laboratories or do they invest in basic life-saving things like first-line treatment of malaria, keeping people on antiretrovirals and so there will be some very difficult tradeoffs there. But I think the thing that's exciting about AI is it allows us to kind of change the equation a bit and maybe get quantum leaps in efficiency and effectiveness. That means we get in a sense more bang for the buck and that's what we got to be... we got to be really focused not on the cool tools so much as where can we really disproportionately have greater impact.
Bill, what about the funding? Is it governments pulling back? What is going on in that environment?
Yeah, almost the top givers to global health have all reduced the amount they give. And so, you know, Gavi raised... which does fundraising every 5 years. They were in Brussels in June for the replenishment hosted by us and Ursula von der Leyen and we raised... we were down over 20%. Global Fund actually we were worried it would be even worse and so things went better than we expected but it's still less money and so the Global Fund board sits there and says okay how much do we cut TB versus malaria versus HIV and that's the situation we find ourselves in and this has real impact. From 2000 to 2024 we had record reductions in childhood death much faster than ever in history. In 2025, for the first time, more children died than the year before. 4.8 million versus 4.6 million. And that's because donors cut money. And some of them cut in a very abrupt unexpected way that really disrupted things like getting malaria chemoprophylaxis out, getting bed nets out. And so we're dependent on these donors. As Peter said, AI is going to help us do more with less. The overhang is not just the development assistance cut but also the indebtedness of the African countries. They are paying more in interest than they paid to run their health care system. So this is the first time that their net interest costs are very high and in the past when this has been the case there have been debt relief and whether the world cares enough and will prioritize that with so many challenges going on that is very difficult. So the story of the miracle of global health and the very positive story of innovation including AI, we have to do a better job getting the voters in these countries to be proud of what they've done and in the face of very tight budgets to maintain what is, you know, less than one... all but the top 10 donors, it's less than 1% of their budget. So in the US, it's well under 1% of the US budget. And yet, we've forgotten how to make that case to get people out to see it, to show them the efficiency that we provide for every dollar that they donate.
Is this something that the private sector can step up around like your partnership today with OpenAI?
Well, yes. In general, the private sector... when you can't just call them up and say, 'Hey, buy measles vaccines to save lives.' They're like busy doing their private sector thing. The tech giants, including OpenAI, do want to devote some of their resources to helping the world at large to show what AI can do. And so they will be partners on a lot of these things. And the computer time that it takes to let the patients in the health care system do all these queries that's going to be provided for free. And so nobody has to think gosh I'm going to have to pay a subscription to get this or that they're going to get some funny ad while they're being told how to deal with malaria. So yes, appealing to the tech sector is important. Every sector you have to think okay what can the agricultural companies do, what can the pharmaceutical companies do and we do a rating, we fund a group that rates the various pharmaceutical companies in terms of their generosity to help out with global health issues. And fortunately people care how they get rated. We'll probably do that for the AI companies at some point just so the ones that are doing a great job get the credit they deserve.
Everybody needs credit. Minister, how have you found the access to funds, the environment? Has it changed? Has it been a challenge? And how are you able to do more with less?
Yes, it's been challenging. And I think one way of doing more with less first is being able to prove value for yourselves but also value for the funding partners because I think to Bill's point, they won't just fund for the sake of funding. There has to be like a win-win type of engagement that you're able to craft and put on the table. But what we've been able to do is really being laser focused with where we deploy the capital that we get, whether it's public or private money that we're able to get. And also to figure out what other programs that could be catalytic in a way that would be able to unlock more funding. So recently last year in Ethiopia we were able to launch the AI scaling hub together with the Gates Foundation and what the scaling hub is looking to do because I think many countries are struggling with this is that many countries have done pilots but what happens, how do you move away from a pilot because a pilot in itself is difficult to even say this is the value, this is the societal impact that we've been able to get and this is how we're able to measure it. And so what we've done through the Gates Foundation is to be able to launch this AI scaling hub where we're picking a few, a handful of use cases that we'll be able to deploy and scale. It does two things. One is that it shows the proof for delivery. Two, you're able to immediately create impact through low-hanging opportunities as opposed to waiting for a big pocket of money that will come for you to sort of fix all the issues and problems that you already have. But what it's also helping is that not every solution has to come from government. So how do we mobilize the industry, the startups that have these ideas and just need a place to test and try some of these solutions and government becomes a test bed or the country becomes a test bed for scaling this. And so I think finding those kind of win-win partnerships is what has made it possible. And obviously as you go along, as you show progress, as you show impact, you're able to mobilize more partnerships and funding.
Right. I mean, Peter, how do you measure it? I mean, the topic du jour on Wall Street is now like what is the return on investment with all of this AI and when is it coming? Is it different than how you previously measured health outcomes and mortality or are we using the same metrics to apply here when it comes to this game-changing tech?
I think fundamentally you want to be coming back to the same metrics. I mean the metrics that we ultimately care about are how many lives have we saved. Cumulative total now is 70 million. And are we reducing infections and ultimately our test is per dollar we have how much progress are we making on those things. And when I look at say the TB screening with digital X-ray, the cost once you've got the digital X-ray machine and they're now these cute little very mobile devices you can take on a Toyota Land Cruiser, the marginal cost of doing an X-ray is extremely small. If you put it through an AI engine and your technology partners have allowed you to do so for free, that's pretty good. And so our cost per high probability case of TB has gone down dramatically. And that to my mind that's a metric I can get really excited about. And so I do think that coming back to the same fundamental metrics and also this comes back to the thing of starting with the problem is we keep looking at what is driving the trends in lives and the trends in infections and that if we work back from that what are the levers that we can pull and which of those levers can we pull harder if we've got an AI engine that is making us smarter or more effective or faster in the way we do it. And look, less money is a bad thing, and has real impact, real impact on lives. The good side of it is that it's a stimulus for us to challenge the way we do things and it's a stimulus for innovation and I think what you are seeing across the sector is actually people looking very hard at how they can do things differently in this environment because we simply have to. If we just do the same thing with less money we're going to end up with more people dying and that is not an answer that we're prepared to live with.
No, Bill. I mean, you mentioned the child mortality numbers which are very depressing. Can that trend change? Can it change as soon as this year even with the lower funding amounts because of what we're talking about?
I don't think the next year or two will be good because we had 10,000 USAID workers that were managing these systems and we're having to change how we do things. Some of those changes in the four or five year time frame actually can be more efficient. We're integrating more into the health system. We're using the new tools. The data tracking is getting a lot better. So the number could go up, maybe five and a half million. It won't go back to 10 million. The goal that the foundation has stated is that during the 20 years that all the money gets spent, we believe we can eradicate some diseases including malaria and polio. And we believe we can get the under five mortality number down, cut it in half again. So that would be below 2.5 million. So we better start to see this upward trend start to bend and come down. But you may not see that for 3 or 4 years.
Three or four years. I was going to ask the timeline.
Yeah. I mean that increase is entirely in Africa. Asia continued to make progress. And in Asia we have countries like India, Vietnam, Indonesia that have grown their economies enough that the amount they need aid has gotten very modest. For example, Gavi has been able to say to India, we'll help you technically with low prices, but you have to fund your work. Global Fund is constantly focusing the money on the countries with the greatest need. They are very analytical about that. And so we have the benefit of those countries freeing up some resources. It means that the countries that we're working in are DRC, Somalia, some of the toughest, but that's our fate is to help those where it's the most challenging.
Well, there's definitely reason for optimism and I think actually this discussion was a really nice start in terms of talking about some of the problems and some of the solutions. So, thank you all for the candid comments and for doing what you do.
Very much appreciated. Thank you guys.
Healthcare workers, doctors and nurses. AI can really be a solution here.
Absolutely. I mean the reverse. We faced this issue with there are well over a million Sudanese refugees in Chad and we had set up mobile clinics with the government of Chad to go into these refugee camps and do screening for TB because frankly whenever you get situations with large numbers of displaced people you end up with TB. And the question is not replacing any radiologists. There were no radiologists. And so if you want the screening to be interpreted, there is no alternative. So in some ways I think one of the reasons this may well take off faster in low middle income countries is because there won't be the resistance from people who say this has taken my job and I don't want to change the way we do things because in fact it's compensating for the fact that those people don't exist. But I want to pick up on the horizontal point that Bill made which is one thing many medical systems in low middle income countries have excelled at is creating paperwork. There is an enormous amount in many healthcare systems of literally paper records with carbon copies and all this kind of stuff and that makes it very difficult to capture because it's on paper. It also just creates an enormous amount of time for overstretched healthcare workers.
Do you mean like records, patient records?
Oh, if you've ever gone to some village and the community health worker will proudly take out a big kind of thing like this and they've carefully written in the name of each patient and what they had and all this kind of stuff and then you think to yourself well a that took quite a long time. But b how is that captured in any kind of national disease surveillance system or so? And the big difference is that now in Rwanda, your community health workers have an app on their phone. It has the details of the patient. They put it in, it's all recorded. They can look at what happened last time. They don't have to go back through the files. And also from a national point of view, that data that's being inputted by the community health worker is being captured. So your community health worker, you've got like 60,000. They become your disease surveillance system, right? And that's incredibly powerful. And then once you've got that kind of level of data capture, you can be applying AI tools for pattern recognition of all sorts of ways of detecting perturbations. You know, this region is seeing something slightly different we didn't expect. So taking paper out of the system, capturing the data, I think that's incredibly powerful.
Can you do that? Can you guys at the Global Fund go and take all those paper stacks and make them?
We're certainly investing in it. I would say that different countries are at very different states in terms of where they are on that pathway. But we do face a challenge, right? There's less money than there was. I mean that is the reality. And a lot of the money that was being invested in things like disease surveillance has stopped.
Well, because major donors around the world have...
Well, it's actually quite broad-based. We've seen quite a lot of donors reducing funding quite significantly. And that means that the investment... and we face a challenge which is that... let me put this in perspective. We've actually just had our eighth replenishment. We've just raised just under 12 billion with some donors still to go. Actually that shows remarkable commitment in this context by the donors who support the Global Fund. But we are going to be allocating less money to countries and they have had sharper reductions from other sources. So countries are going to face some very difficult choices. Do they invest in underlying system things like disease surveillance and laboratories or do they invest in basic life-saving things like first-line treatment of malaria, keeping people on antiretrovirals and so there will be some very difficult tradeoffs there. But I think the thing that's exciting about AI is it allows us to kind of change the equation a bit and maybe get quantum leaps in efficiency and effectiveness. That means we get in a sense more bang for the buck and that's what we got to be... we got to be really focused not on the cool tools so much as where can we really disproportionately have greater impact.
Bill, what about the funding? Is it governments pulling back? What is going on in that environment?
Yeah, almost the top givers to global health have all reduced the amount they give. And so, you know, Gavi raised... which does fundraising every 5 years. They were in Brussels in June for the replenishment hosted by us and Ursula von der Leyen and we raised... we were down over 20%. Global Fund actually we were worried it would be even worse and so things went better than we expected but it's still less money and so the Global Fund board sits there and says okay how much do we cut TB versus malaria versus HIV and that's the situation we find ourselves in and this has real impact. From 2000 to 2024 we had record reductions in childhood death much faster than ever in history. In 2025, for the first time, more children died than the year before. 4.8 million versus 4.6 million. And that's because donors cut money. And some of them cut in a very abrupt unexpected way that really disrupted things like getting malaria chemoprophylaxis out, getting bed nets out. And so we're dependent on these donors. As Peter said, AI is going to help us do more with less. The overhang is not just the development assistance cut but also the indebtedness of the African countries. They are paying more in interest than they paid to run their health care system. So this is the first time that their net interest costs are very high and in the past when this has been the case there have been debt relief and whether the world cares enough and will prioritize that with so many challenges going on that is very difficult. So the story of the miracle of global health and the very positive story of innovation including AI, we have to do a better job getting the voters in these countries to be proud of what they've done and in the face of very tight budgets to maintain what is, you know, less than one... all but the top 10 donors, it's less than 1% of their budget. So in the US, it's well under 1% of the US budget. And yet, we've forgotten how to make that case to get people out to see it, to show them the efficiency that we provide for every dollar that they donate.
Is this something that the private sector can step up around like your partnership today with OpenAI?
Well, yes. In general, the private sector... when you can't just call them up and say, 'Hey, buy measles vaccines to save lives.' They're like busy doing their private sector thing. The tech giants, including OpenAI, do want to devote some of their resources to helping the world at large to show what AI can do. And so they will be partners on a lot of these things. And the computer time that it takes to let the patients in the health care system do all these queries that's going to be provided for free. And so nobody has to think gosh I'm going to have to pay a subscription to get this or that they're going to get some funny ad while they're being told how to deal with malaria. So yes, appealing to the tech sector is important. Every sector you have to think okay what can the agricultural companies do, what can the pharmaceutical companies do and we do a rating, we fund a group that rates the various pharmaceutical companies in terms of their generosity to help out with global health issues. And fortunately people care how they get rated. We'll probably do that for the AI companies at some point just so the ones that are doing a great job get the credit they deserve.
Everybody needs credit. Minister, how have you found the access to funds, the environment? Has it changed? Has it been a challenge? And how are you able to do more with less?
Yes, it's been challenging. And I think one way of doing more with less first is being able to prove value for yourselves but also value for the funding partners because I think to Bill's point, they won't just fund for the sake of funding. There has to be like a win-win type of engagement that you're able to craft and put on the table. But what we've been able to do is really being laser focused with where we deploy the capital that we get, whether it's public or private money that we're able to get. And also to figure out what other programs that could be catalytic in a way that would be able to unlock more funding. So recently last year in Ethiopia we were able to launch the AI scaling hub together with the Gates Foundation and what the scaling hub is looking to do because I think many countries are struggling with this is that many countries have done pilots but what happens, how do you move away from a pilot because a pilot in itself is difficult to even say this is the value, this is the societal impact that we've been able to get and this is how we're able to measure it. And so what we've done through the Gates Foundation is to be able to launch this AI scaling hub where we're picking a few, a handful of use cases that we'll be able to deploy and scale. It does two things. One is that it shows the proof for delivery. Two, you're able to immediately create impact through low-hanging opportunities as opposed to waiting for a big pocket of money that will come for you to sort of fix all the issues and problems that you already have. But what it's also helping is that not every solution has to come from government. So how do we mobilize the industry, the startups that have these ideas and just need a place to test and try some of these solutions and government becomes a test bed or the country becomes a test bed for scaling this. And so I think finding those kind of win-win partnerships is what has made it possible. And obviously as you go along, as you show progress, as you show impact, you're able to mobilize more partnerships and funding.
Right. I mean, Peter, how do you measure it? I mean, the topic du jour on Wall Street is now like what is the return on investment with all of this AI and when is it coming? Is it different than how you previously measured health outcomes and mortality or are we using the same metrics to apply here when it comes to this game-changing tech?
I think fundamentally you want to be coming back to the same metrics. I mean the metrics that we ultimately care about are how many lives have we saved. Cumulative total now is 70 million. And are we reducing infections and ultimately our test is per dollar we have how much progress are we making on those things. And when I look at say the TB screening with digital X-ray, the cost once you've got the digital X-ray machine and they're now these cute little very mobile devices you can take on a Toyota Land Cruiser, the marginal cost of doing an X-ray is extremely small. If you put it through an AI engine and your technology partners have allowed you to do so for free, that's pretty good. And so our cost per high probability case of TB has gone down dramatically. And that to my mind that's a metric I can get really excited about. And so I do think that coming back to the same fundamental metrics and also this comes back to the thing of starting with the problem is we keep looking at what is driving the trends in lives and the trends in infections and that if we work back from that what are the levers that we can pull and which of those levers can we pull harder if we've got an AI engine that is making us smarter or more effective or faster in the way we do it. And look, less money is a bad thing, and has real impact, real impact on lives. The good side of it is that it's a stimulus for us to challenge the way we do things and it's a stimulus for innovation and I think what you are seeing across the sector is actually people looking very hard at how they can do things differently in this environment because we simply have to. If we just do the same thing with less money we're going to end up with more people dying and that is not an answer that we're prepared to live with.
No, Bill. I mean, you mentioned the child mortality numbers which are very depressing. Can that trend change? Can it change as soon as this year even with the lower funding amounts because of what we're talking about?
I don't think the next year or two will be good because we had 10,000 USAID workers that were managing these systems and we're having to change how we do things. Some of those changes in the four or five year time frame actually can be more efficient. We're integrating more into the health system. We're using the new tools. The data tracking is getting a lot better. So the number could go up, maybe five and a half million. It won't go back to 10 million. The goal that the foundation has stated is that during the 20 years that all the money gets spent, we believe we can eradicate some diseases including malaria and polio. And we believe we can get the under five mortality number down, cut it in half again. So that would be below 2.5 million. So we better start to see this upward trend start to bend and come down. But you may not see that for 3 or 4 years.
Three or four years. I was going to ask the timeline.
Yeah. I mean that increase is entirely in Africa. Asia continued to make progress. And in Asia we have countries like India, Vietnam, Indonesia that have grown their economies enough that the amount they need aid has gotten very modest. For example, Gavi has been able to say to India, we'll help you technically with low prices, but you have to fund your work. Global Fund is constantly focusing the money on the countries with the greatest need. They are very analytical about that. And so we have the benefit of those countries freeing up some resources. It means that the countries that we're working in are DRC, Somalia, some of the toughest, but that's our fate is to help those where it's the most challenging.
Well, there's definitely reason for optimism and I think actually this discussion was a really nice start in terms of talking about some of the problems and some of the solutions. So, thank you all for the candid comments and for doing what you do.
Very much appreciated. Thank you guys.
Healthcare workers, doctors and nurses. AI can really be a solution here.
Absolutely. I mean the reverse. We faced this issue with there are well over a million Sudanese refugees in Chad and we had set up mobile clinics with the government of Chad to go into these refugee camps and do screening for TB because frankly whenever you get situations with large numbers of displaced people you end up with TB. And the question is not replacing any radiologists. There were no radiologists. And so if you want the screening to be interpreted, there is no alternative. So in some ways I think one of the reasons this may well take off faster in low middle income countries is because there won't be the resistance from people who say this has taken my job and I don't want to change the way we do things because in fact it's compensating for the fact that those people don't exist. But I want to pick up on the horizontal point that Bill made which is one thing many medical systems in low middle income countries have excelled at is creating paperwork. There is an enormous amount in many healthcare systems of literally paper records with carbon copies and all this kind of stuff and that makes it very difficult to capture because it's on paper. It also just creates an enormous amount of time for overstretched healthcare workers.
Do you mean like records, patient records?
Oh, if you've ever gone to some village and the community health worker will proudly take out a big kind of thing like this and they've carefully written in the name of each patient and what they had and all this kind of stuff and then you think to yourself well a that took quite a long time. But b how is that captured in any kind of national disease surveillance system or so? And the big difference is that now in Rwanda, your community health workers have an app on their phone. It has the details of the patient. They put it in, it's all recorded. They can look at what happened last time. They don't have to go back through the files. And also from a national point of view, that data that's being inputted by the community health worker is being captured. So your community health worker, you've got like 60,000. They become your disease surveillance system, right? And that's incredibly powerful. And then once you've got that kind of level of data capture, you can be applying AI tools for pattern recognition of all sorts of ways of detecting perturbations. You know, this region is seeing something slightly different we didn't expect. So taking paper out of the system, capturing the data, I think that's incredibly powerful.
Can you do that? Can you guys at the Global Fund go and take all those paper stacks and make them?
We're certainly investing in it. I would say that different countries are at very different states in terms of where they are on that pathway. But we do face a challenge, right? There's less money than there was. I mean that is the reality. And a lot of the money that was being invested in things like disease surveillance has stopped.
