AI, Data Governance, and Innovation for Development

Summary

This discussion focused on AI, data governance, and innovation for development in West Africa, exploring challenges and opportunities in the region. Panelists emphasized the importance of creating enabling regulatory environments through co-creation and harmonized policies. They highlighted the need for problem-driven research and collaboration between academia, industry, and local communities to develop AI solutions tailored to regional needs.

A key challenge identified was the lack of locally relevant datasets, with panelists stressing the importance of developing inclusive, culturally appropriate data. They discussed the need for ethical data collection practices, responsible use guidelines, and centralized data repositories. The digital divide was addressed, with suggestions for innovative funding models and regulatory frameworks to expand connectivity to rural areas.

The discussion touched on the responsible use of AI in education, with panelists advocating for frameworks that allow AI use while ensuring students demonstrate understanding. They emphasized the importance of digital literacy and adapting to evolving technologies to remain competitive globally.

Participants also explored the potential for public-private partnerships and the role of universities in fostering AI innovation. The need for high-performance computing infrastructure and funding for data collection and curation was highlighted. Overall, the discussion underscored the transformative potential of AI for West Africa’s development, while emphasizing the importance of addressing challenges around data, infrastructure, and capacity building.

Keypoints

Major discussion points:

– Creating enabling regulatory environments for AI innovation through co-creation, collaboration, and harmonized policies

– Developing indigenous AI solutions and datasets tailored to local African problems and contexts

– Addressing connectivity and digital divide issues to ensure equitable AI adoption, especially in rural areas

– Responsible use of AI in education and research, balancing innovation with ethics and learning

– Need for funding, infrastructure, and training to support AI development and data governance in Africa

The overall purpose of the discussion was to explore how West African countries can leverage AI, data governance, and innovation for sustainable development while addressing challenges unique to the region.

The tone of the discussion was largely optimistic and solution-oriented. Speakers acknowledged significant challenges but focused on practical ways to overcome them through collaboration, policy changes, and capacity building. The tone remained constructive throughout, with an emphasis on seizing opportunities while being mindful of risks.

Speakers

– Sade Dada: Head of Public Policy, Anglophone West Africa, and Connectivity Innovation Policy Lead at META

– Wale Adedokun: Professor of Artificial Intelligence and Computer Science at Amadou Bello University, Zaria; Deputy Director of the National Center for International Online Institutes

– Martha Omoekpen Alade: Executive Director of Women in Technology in Nigeria (WITN); Moderator of the session

Additional speakers:

– Salieu Kanu Mansaray: Chief Information Security Officer at Mammoth Sierra Leone (mentioned but did not participate)

– Isola Olorunisomo: Director of Research and Strategy at Nigeria Data Protection Commission (mentioned but did not participate)

AI, Data Governance, and Innovation for Development in West Africa: A Comprehensive Discussion

This report summarises a panel discussion on artificial intelligence (AI), data governance, and innovation for development in West Africa. The discussion explored challenges and opportunities in the region, with a focus on creating enabling regulatory environments, developing indigenous AI solutions, addressing connectivity issues, and promoting responsible AI use in education and research.

Participants:

– Sade Dada: Head of Public Policy, Anglophone West Africa, and Connectivity Innovation Policy Lead at META

– Wale Adedokun: Professor of AI and Computer Science at Ahmadu Bello University, Zaria; Deputy Director of the National Center for International Online Institutes

– Martha Omoekpen Alade: Executive Director of Women in Technology in Nigeria (WITN); Moderator of the session

Key Themes and Discussions:

1. Creating an Enabling Regulatory Environment

The panellists emphasised the importance of developing regulatory frameworks that foster AI innovation while addressing potential risks. Sade Dada highlighted the value of co-creation in this process, advocating for collaboration between various sectors impacting regulatory agendas and innovation. She stressed the need to harmonise existing regulations to create a cohesive framework for AI development.

Professor Wale Adedokun emphasised the importance of problem-driven research focused on local challenges. He advocated for engaging communities in data collection and AI solution development, ensuring that innovations address real-world problems faced by West African populations.

2. Connectivity and Digital Divide

Addressing the digital divide was identified as crucial for equitable AI adoption across West Africa. Sade Dada proposed innovative funding models for rural infrastructure projects and suggested reconsidering spectrum allocation policies to enable novel connectivity solutions. She also highlighted the importance of addressing digital literacy and affordability to increase AI adoption.

3. Developing Inclusive and Culturally Relevant Datasets

A key challenge identified was the lack of locally relevant datasets for AI development in West Africa. Professor Adedokun stressed the importance of creating local datasets in indigenous languages, covering various sectors such as agriculture, education, and health. He advocated for establishing ethical data collection and curation practices, as well as developing centralised repositories for sharing local data.

The discussion highlighted the need for funding and infrastructure to support data collection efforts. Panellists emphasized the importance of ethical data practices and responsible use. The creation of a central repository for localized data was suggested as a way to facilitate AI development in the region.

An audience member mentioned the “Wikis for Seniors” project as an example of content creation for AI, highlighting the potential for community-driven data generation.

4. Responsible Use of AI in Education

The discussion touched on the responsible use of AI in education, with panellists addressing concerns raised by students about professors warning against using AI for assignments. Professor Adedokun suggested developing guidelines for appropriate AI use in academic settings, including asking students to provide the prompts they used with AI tools.

Sade Dada emphasised the need to train professors on understanding and leveraging AI tools. She framed AI adoption as crucial for remaining competitive on the global stage, warning against the potential consequences of not embracing AI in education and research.

Martha Omoekpen Alade highlighted the importance of balancing AI assistance with original thinking and work. There was a consensus among the speakers on embracing AI tools in education while maintaining academic integrity.

5. Infrastructure and Funding for AI Development

The panellists discussed the need for high-performance computing infrastructure and funding for data collection and curation. They explored the potential for public-private partnerships and the role of universities in fostering AI innovation.

Professor Adedokun shared an example of undergraduate students developing a Crop Disease Detector, demonstrating the practical application of AI to address local agricultural challenges. This project showcased how AI solutions can be developed to solve specific regional problems.

Future Directions:

Martha Omoekpen Alade suggested organizing training on data governance for researchers across Nigeria, highlighting the need for capacity building in this area. The development of a framework for responsible use of AI in tertiary institutions was mentioned as already in progress.

Conclusion:

The discussion underscored the transformative potential of AI for West Africa’s development while emphasising the importance of addressing challenges around data, infrastructure, and capacity building. Key focus areas included:

1. Developing locally relevant datasets and establishing ethical data collection practices

2. Creating enabling regulatory environments that foster innovation

3. Addressing the digital divide through improved connectivity and digital literacy

4. Promoting responsible AI use in education while ensuring global competitiveness

5. Investing in infrastructure and funding for AI research and development

The overall tone was optimistic and solution-oriented, with speakers focusing on practical ways to overcome obstacles through collaboration, policy changes, and capacity building. As the region moves forward, balancing aspirational AI development with solving fundamental issues will be crucial for realizing the full potential of AI in driving sustainable development across West Africa.

Martha Omoekpen Alade: Good afternoon, everyone, and a very warm welcome to the West Africa Internet Governance Forum 2025 Closing Plenary, titled AI, Data Governance and Innovation for Development. My name is Martha Omoekpen-Alade, the Executive Director of Women in Technology in Nigeria, WITN. And I am truly honoured and excited to moderate this crucial session today. I would like to extend a special warm welcome and heartfelt thanks to our esteemed panelists. Thank you for being panelists. It’s a privilege to have such distinguished group with us. Please join me in welcoming Sade Dada, the Head of Public Policy, Anglophone West Africa, and Connectivity Innovation Policy Lead at META. Sade, you are welcome. A round of applause for her, please. Then we have with us, joining us online, Wale Adedokun, is a Professor of Artificial Intelligence and Computer Science at Amadou Bello University, Zaria. We have also with us Salieu Kanu Mansaray, the Chief Information Security Officer at Mammoth Sierra Leone. He’s joining us online as well. And finally, we have Isola Olorunisomo, the Director of Research and Strategy at Nigeria Data Protection Commission. I don’t know if he’s here with us. Okay, he’s not here. Hopefully, he will join us. Okay. So to set the stage for discussion, we’ll start by having each of our panelists briefly introduce themselves and their work. So let’s start with you, Sade. Could you tell us a bit about yourself and what you do at META?

Sade Dada: Thank you so much, Martha. Thank you to the organized important dialogue. The sessions have been really great over the last couple of days. So as the introduction said, I lead META’s public policy work in Anglophone West Africa primarily. And I also work across Africa on issues related to to connectivity and innovation, emerging technologies like AI, which is the subject of the session today. So my job entails interacting with government, civil society organizations, other industry players, and coming up with solutions to help create enabling environments for innovation to thrive in various countries across the continent.

Martha Omoekpen Alade: Thank you very much Sade. Yes, thanks for that brief introduction. So let’s go over to Isola is not here yet. Okay. What about Professor Wale Adedokun? Is he online? Okay. Professor Prof. Yes, please just introduce yourself briefly and tell us a bit about your work as a prof.

Wale Adedokun: Good afternoon. Good afternoon, everybody. I’m very glad to be here. I will seek your indulgence to put up the video because network is not very good here, but so that my voice can be very clear. Thank you. Okay. So my name is Wale Adedokun. I work with Amadu Bilo University, Zaria, Department of Computer Engineering. I also serve as the Deputy Director of the National Center for International Online Institutes, which is a UNESCO project. Basically what we try to do is digital transformation, and one of the major things that UNESCO is promoting now is the use of artificial intelligence and development of artificial intelligence in HAIs in Nigeria. So currently we are doing quite a bit on developing training modules. Using Artificial Intelligence Virtual Reality, we already have a laboratory that trains graduates and undergraduate students on the use of artificial intelligence and how to actually develop, do a lot of research in that area. As we continue the discussion, I will probably talk about some of the projects that we are working on now, some of the docs that we have developed using artificial intelligence in collaboration with the industry. So, basically, I think I should stop now and we continue the discussion. Thank you.

Martha Omoekpen Alade: Thank you very, very much. So, let’s quickly go over to Salieu. I don’t know if he’s online with us. Salieu Mansaray, the Chief Information Security Officer at Mammoth, Sierra Leone. Could you check online if he is? Okay, he’s online. While we wait for him, let’s go on with Sade. So, Sade, given your work at the intersection of public policy and innovation at META, how can West African countries create regulatory environments that enable AI innovation while safeguarding against misuse, especially considering our diverse cultural and socioeconomic realities?

Sade Dada: Thank you. I believe that is a really, really important question. I mean, the whole purpose of us gathering here as West Africa IGF is to to look at how we can work together to work on policies related to internet governance and how different players in the ecosystem should come together to make that happen. When it comes to an enabling regulatory environment, what I’ve seen in the course of my work that is very useful is when we focus on co-creation. And what do I mean by that? That means that you collaborate with various parties that impact a particular regulatory agenda or a particular field or innovation sector. So we’re talking about AI now. When we want to have a regulatory environment that enables innovation around AI, but we should not do that in a silo. We shouldn’t come up with a policy or strategy in a silo. What’s most useful and what I’ve seen work in certain countries that I’ve operated in is when, for instance, in Nigeria, we came up with the national AI strategy in a way that actually included so many different sectors. Academia was involved. The private sector was involved, different government agencies, legislators. It was very, very much a co-creation environment. And that really led to bringing out insights that we otherwise would not have had to influence the policy itself, I mean, the strategy itself. And so what we saw through that process is that people brought expertise around things like safety and ethics and cultural nuances. But then you had people who were on the developer side who understood how the technology itself works and can use that to influence the kind of policies that would be necessary to enable that, but also to have checks and balances. And all of those things came out. because we did it in a co-creation type of way. And so kudos to the ministry, the Nigerian Ministry of Communications, Innovation and Digital Economy. I think that was the right approach to take. The other thing that is extremely important is that we harmonize review of regulations. So many times when we see a new technology that comes into play, we think we automatically have to create a new regulation or a new policy that’s specific to that particular innovation. Whereas if we keep doing it that way, we’re always gonna be playing catch up. So rather than doing that for every single new technology that actually relates to the technology and the way we use it and what updates may be necessary to that, if we don’t do it that way, we run into a situation where we have a number of different regs that can apply to a particular technology and some of them are not in agreement with each other. And so that then creates an environment that is actually very harmful to promoting innovation. And so we see in some countries where they actually undertake that regulatory review as a part of the process of figuring out how to create a strategy around AI. So they look at their data protection laws, they look at laws around connectivity and broadband deployment. They look at regulations around online safety and you have to really take all of those elements into account in order to have an environment from the regulatory perspective that is conducive. The last point that I’ll make on this is that you have to carry people along to be able to understand what those laws and regulations mean, right? After you’ve enacted them, because a lot of times people who work in the innovation space may not actually understand. They’ll know what compliance looks like, what safeguards they have in place… are supposed to before. For a company like ours at Metta, we do things like we come up with responsible use guides for our AI models where we teach people these sort of things. So if they’re going to utilize our large language model, for instance, they need to understand the parameters around doing that. And we give expertise around data protection, around online safety, around how do you do cybersecurity correctly. And all of those things are what we should keep in mind when it comes to this in regulatory environment that we’re trying to foster.

Martha Omoekpen Alade: Thank you very much, Sade. I was just smiling. For about two, three years, I and Sade, we’ve worked closely in some policy groups. So I kept asking myself, Sade is a busy person, you know, working across Africa. How come she has the time to join us, you know, in those small policy groups? But now I think I understand because of your commitment to co-creation and you really want to be on ground. Thank you very much, Sade, for that. So let me go quickly to Prof again. So Prof Adedokun, as a leading academic in AI, how can universities in most or in West Africa in general move beyond theory to develop indigenous AI solutions tailored to local problems in sectors like agriculture, education and health care? Bob, you got me right?

Wale Adedokun: Yes, ma’am. Yeah. Thank you. Okay. Thank you very much. I will just itemize some of my thoughts very quickly. The very first thing that I think we need to promote now, as the academia needs to promote, is to begin to develop problem-driven research. I mean, if we begin to focus our researches on real-world challenges, I mean situations, challenges around us, local communities, begin to look at local communities, what are the issues they have, how can we use AI to solve these problems, this will help not only to promote research, but it will have a direct effect on the communities themselves. So, for example, I will tell you one of the things that we are doing. Currently now, as we speak, some of our undergraduate students are in China, representing Africa, with a product that they just developed, and the product is called Crop Disease Detector. So, what that product really does is, when you see a disease on a crop in the farm, the product is going to detect that particular disease and probably profile possible solution to those diseases. So, this product was developed basically collecting local data from the farms and from our agri-complex within the university. Now, that product won the competition in Africa, and currently they are in China, because the competition is promoted by Huawei. So, when you develop such products using AI, it has a direct impact on the local communities. So, that’s number one. Two, Sade also mentioned collaboration, community engagement. When you want to develop products that you want the community to appreciate, you need to engage them. You need to be able to talk to people when you are collecting your data. One of the major problems we have really with AI, it’s not that Nigerian or West Africa researchers are not doing a lot, but one of the major problems we have is data. Most of the AI products you want to develop, you will discover that we use data from the West. And at times, they don’t really speak to our contemporary issues in West Africa. So community engagement is very important. A lot is happening from the Ministry of Communication and Digital Economy, especially when it comes to innovation labs. And we have also seen this gap even in the universities. In my department, we just registered a hub. We called it Concept to Product Hub. So basically, what we want to do is, the researchers that we have, we want to take it into the hub and see how we can produce them into products, products that will make direct impact on our immediate community. Also, because of probably funds and some other issues, collaboration, like Sade said, with NGOs, with governments, departments, becomes very, very important. When we try to do all these, we begin to have direct impact on our community. Student engagement is very important. Sustainable funding is very important. When you do something and there is no fund, most of the researchers will just end up as just, you know, academic work, basically. But when you get the industry involved in the researches you do, They pick those products because they directly affect the communities. They pick those products or the researchers and they go back and turn them into products that become very useful to the immediate community. Thank you.

Martha Omoekpen Alade: Thank you very much, Prof. I was listening with keen interest when you mentioned that you now have a hub where your research work can really be put to the field. Not just, lots of researchers all over Nigeria that are just in the book. So I am really excited to hear that ABU is doing something different. Thank you very, very much. I don’t know if Salieu is online. Mansaray. So we are still expecting Salieu, Mansaray and Olorunisomo from NDPC. Is any of them online? Okay. All right. So I have to come back to you again, Sade. So Sade, once more, again, connectivity remains a foundational issue for equitable AI adoption in this region in West Africa. So what are some scalable policy or public-private partnership models that could breach this digital divide and ensure AI innovations reach rural and underserved communities? So we are talking about those communities that are not really literate. They use Facebook, they use Instagram, but not as much as those in the urban areas. So over to you.

Sade Dada: So, you know, getting to these areas is really, really complicated, very, very challenging, and it’s because of the market dynamics. So people, the companies that do telecom infrastructure have to see that they will have a return on their investment. And that’s what makes it quite difficult to go into areas that are rural, because the cost of deploying is not matching how much they can actually make from doing so. So what we’ve seen is that you have to come up with very unique funding models for doing those types of infrastructure projects. For instance, as a company, we had been working on what’s called network as a service for many years, where you have an infrastructure provider, that will be the anchor for shared infrastructure, both passive and active infrastructure. And that requires a very unique regulatory allowance to be able to do that type of infrastructure sharing. But it’s one of the things that we’ve seen actually work to get to rural areas. And whenever you see that type of collaboration that happens between an infrastructure provider, maybe a telecom operator, you might have some investment bank or another private sector player like a tech company all put funds together to come up with a particular solution that involves sharing. We see that that has actually worked in countries that have difficulty reaching those rural areas. But it requires forward-thinking regulation that actually allows for that type of sharing. We also look at things like having access to spectrum that is not limited to just providing licenses to operators. Operators, again, have a very unique or a very specific way that they are able to make money from doing infrastructure deployment. But we see in situations, for instance, if you look at the six gigahertz band of spectrum, there’s been ongoing debate about how to use that particular spectrum band. We’ve seen countries where they opened up either the full band or a portion of the band for unlicensed Wi-Fi use. Why am I bringing it up in this scenario? Because we actually see use cases where Wi-Fi was paired with satellite services to provide connectivity in rural areas. I’m thinking of a particular project that I saw happen in DRC utilizing that spectrum band. We have to think outside of the box when it comes to reaching those unique funding models, having a regulatory framework that enables new market entrants to have new solutions. For instance, community networks, how do we enable them to be able to provide connectivity? Unless we’re doing that, we’re not going to have everybody be able to reap the benefits of AI solutions that are coming out today. The last thing I would say is that connectivity coverage, access to broadband, is not limited to having the infrastructure available. You have to remember that you still need people to actually get online once the infrastructure is available. There’s been studies for instance that the Inclusive Internet Index did over a few years. It’s a study done by an organization called The Economist where they tracked the level of broadband coverage across multiple countries. Many of our countries are part of this. And what they found over the years is that even though coverage had increased substantially, actual adoption was not increasing at the same rate. And what that tells me is that we have to also look at policies that are geared towards digital literacy, about how to use online tools but also use it safely. You have to look at issues around affordability. You have to look at issues around the difference in gender norms when it comes to getting online. All of those things are part of the connectivity process and bringing more people online so that they can reap the benefits of these various new solutions that are coming out like AI.

Martha Omoekpen Alade: Thank you very much, Sade. You are just making me smile. Whenever I hear people talk about reaching the marginalized and with concrete solutions, it gives me joy. Thank you so much. I do appreciate that. So I don’t know if Selu is online, Mansareh, is he online? Or Prof, no, the other speaker, I think, no other person. Okay, so another person online. Okay, let’s go back to Prof. After Prof speaks, then we will open the floor for contributions and questions. So Prof, I have one more question for you, and this really touches me deep in my heart because it’s something if addressed. will transform the AI landscape of not just West Africa, Africa as a whole. So we know that AI bias often stems from the datasets used to train models. So Prof, in your experience, what role can African researchers play in developing more inclusive, culturally relevant datasets? And what infrastructure is needed to support this?

Wale Adedokun: Once again, thank you. Like I mentioned earlier, you also alluded to data is AI, or AI is data. Basically, when you talk about developing any AI solution, it really has to do with the data you use to train, the data you use to test. That basically determines the efficiency or effectiveness of that AI solution. So for researchers, sincerely, I would say that researchers are trying their best. But I would match everything that has to do with data, probably three things that have to do with data, I would match them into one. Local dataset development, we need to begin to look at, if we want our researchers to be very relevant, start to develop local datasets. So let’s look at datasets, whether it is in agriculture, it’s in education, it’s in health. And not just create these datasets in English language, we also need to begin to see how we can create datasets in our local languages. So that when you develop a model, it speaks directly to… the community because this model can understand and speak the local languages. This is easy to say, it’s not very easy to do. When you also talk about data, you also need to begin to understand the ethics behind data collection and curation. If you want to collect medical data, basically, you want to develop an AI model for probably a disease or something, like one of my M.Sc students was trying to do an AI model that depicts the emotion of a pregnant woman just from the voice of the woman. Now, how do you collect this data? How do you store this data? Most pregnant women, when you get to the hospital, they don’t want to talk to you. Now, even when they talk to you, the ethics behind using the data is something that everybody needs to begin to learn. When you are doing research, it has to be that there are ethics. We need to begin to learn ethical data practices, responsible use of this data itself. Now, we need to now begin to think of how we have probably a central repository localized, whether it is national, whether it is regional, where this data can be kept and shared. All the data that we use today, they are kept somewhere, they are shared, but they are not local data. One of our students at some point was trying to develop an AI model that predicts, identifies skin lesions. Now, most of the data that we have to use to train that model are data from the West. The skin lesion of a white person is not the same with the skin lesion of a black person. Now, if you want to collect this data, even from the medical sector, it becomes a lot of problems. So, we need a lot of advocacy. We need people to understand that we must grow, develop things that speak to our problems directly. We must be able to collect all this data with a clear understanding that this data will be used responsibly. This data will be available to the community or to a committee of researchers. So, instead of downloading data from the West, you have a repository that speaks directly to our needs. People can pull out data, well-curated data, they can pull them out and use this data to develop their AI models. Infrastructure, like I said, you need data storage, large data storage management system. At the same time, we need access to high-performing computing clusters. Because this data, this development needs a lot of high-performing systems. So, we need high-performing computing systems. Clusters all over the place. Nigeria is a very large place. I don’t know how many clusters we need, but even for Nigeria alone, we need a lot. Grants, funding, you know, for data collection. Data collection is a major thing and it needs a lot of… Some people you want to collect data from, you need to… Sade Dada, Wale Adedokun, Isola Olorunisomo, Wale Adedokun, Isola Olorunisomo, Wale Adedokun, a very good policy framework that it helps a lot, and it’s going to go a long way to encourage not only the researchers, but even the African community to begin to do things that speak to our needs directly. Thank you.

Martha Omoekpen Alade: Thank you so much, Prof. That was excellent. In fact, you mentioned funding. Collecting data is no child’s play, so grants are crucial, funding important, we need infrastructure, and then we need to advocate for open data. Because there are lots of data in silos here and there, so we need concerted effort, policy, support, and even not just advocacy, enlightening researchers on how they can find good repositories to put in data and make it open, so we need a lot of education as well. Thank you so much, Prof. So for the last time, let’s check if we have any of the speakers online to join us. Is there anyone? Seliu? None. Okay, so well done to our speakers. So I think it’s safe to say we can open the floor for questions and contributions. So if you have questions, or you want to make a little contribution briefly, can I see your hands up? Okay, please pass the mic.

Audience: Thank you. I’m by the name of Comrade Abdelrahman Ousmane Fatika from the prestigious University Ahmed Belur University Zaria, a student of Sociology. So my question goes to the professor and the panelists is that, Ma, we have this problem on ground. When you are given assignment or you want to make some research, the professors or the lecturers will be warning you that don’t use AI to make my research. They will be warning you that don’t use AI. So Ma, this challenge that we are having in school, that is to say we have to adopt, we have to keep AI aside and be using other side or you can use the AI again. So there is the challenge, how can we overcome this challenge? Thank you.

Martha Omoekpen Alade: I think, do we take all the questions or will Prof answer as they ask? Let’s take all the questions. Prof, I hope you are taking notes of the questions.

Wale Adedokun: Yes, I am.

Martha Omoekpen Alade: Because that question was directed to you. So let’s take all the questions then, Prof will answer once and Sade as well. Okay, next question.

Audience: My name is Jose from Ghana. My question is to all the panelists. I’d like to know, I mean, AI has come as a real game changer and there seems to be a fundamental issue that is existing or is there a growing AI digital divide? And I’d like to know, where can we find a balance in we being aspirational and also solving the fundamental issues? Precisely, yeah.

Martha Omoekpen Alade: Okay, the third person.

Audience: Thank you. I’m not sure my own is a question. You see, every time I hear our community talk about AI, I want us to first of all understand that AI is garbage in garbage out. If you ask AI, just do AI. It’s on to me, it puts there, the name of my grandmother is so-so-so. Then let another person ask tomorrow, what’s Dr. Omoekpen Alade’s grandmother’s name? AI will answer you. So when we say that the lecturer tells you that you shouldn’t answer with AI, it’s because we are becoming too dependent. And we see more of copy and paste. Even though AI has answers for us, can we be so literate a little to be able to modify it and tweak it to what is useful to what we need it for? So this is I’m sure that’s probably why the lecturer says that. And that’s by the way, what that means to us is that at this side of the world, we need a lot of more content developers, people who should be, there is this Wikis for Seniors. When I was having my section, somebody introduced himself as a Wikipedia. And I actually said, please let’s talk because I am on that platform, Wikis for Seniors. And what we are doing there is that senior citizens, those who are retired, who still have indigenous knowledge, who knows much more, they can tell the stories, the histories of so many things. They are on that platform, and they are putting content into Wikipedia. And so when we talk about AI, and these issues about data collection, Mata said, we need funding, serious funding, even for us to sit down in this room and say what’s happened

Martha Omoekpen Alade: So, Martha Omoekpen Alade has actually helped us to answer one of the questions, right? Okay, but Prof, we still, you know, shed more light to that question. Any other question? Okay, so we are good. All right. So, Prof, could you take the first, respond to the first question, where the students are being restricted in the use of AI in the university. Over to you, Prof.

Wale Adedokun: Okay, thank you. We have talked about this, and sometimes last year, we had a national discourse on developing a framework for the use of AI in our HAIs. And basically, this is one of the reasons why we started doing that. For us, we already have a framework, but we know we can’t localize it just to our university. So, we had a national dialogue, followed up with several online sessions. At the end of the day, we have a framework that is already with the Federal Ministry of Education and National University Commission on responsible use of AI in tertiary institutions. Now, we have gotten to a point that really, it may be difficult to stop students, researchers from using AI. And for me, if I give you an assignment, I will not stop you from using AI. What I will ask you to do is, when you are submitting your assignment, if you use AI, make sure that you send me the prompt that you give to AI. When I see your prompt, it gives me an understanding whether you know what you are doing. or you don’t know what you are doing. Whether you are depending on AI to give you all the solutions. Being, like something we call prompt engineering, being able to give AI the right prompt alone gives me an understanding of whether you know what you are doing or you don’t know what you are doing. So, and that’s the place to start. Eventually, we are getting to a point that we’ll be able to say, just like pressure research, we’ll be able to say when you are submitting an assignment, your AI should not be more than probably 20%, 30%, 40%. Everything should not be something that was generated by a system and given to you. I’m sure you have seen somewhere online that somebody was graduating and he said, I thank Chat GPT for doing my project for me. I thank this. So, basically, they do nothing. They just go to Chat GPT. He gives them everything and they give it to the lecturer. They say they have done a project. So, we need to also understand that as a student, if you have come to learn, you should be able to think by yourself. If you are able to think, you generate a prompt. Your prompt that you are giving to AI will tell me that you are thinking or you are not thinking. And when AI generates a response to your prompt, when you give it to me, I will know whether you have done anything to it or you are just giving me what AI is giving you verbatim. But true to it, we have gotten to a point that we cannot stop the use of AI. Basically, that’s why AI is there. We only need to regulate the use and ensure that it is used respectfully.

Martha Omoekpen Alade: Thank you very much, Prof. I really like that idea of you asking the students what prompts they use. And to the young man who asked that question, you know, there are responsible ways you could actually use AI to do whatever work you want to do. You can use that to generate ideas and still do your work and do it perfectly well. So I don’t think there’s anything wrong with that. But where it becomes really bad is when you just copy and paste. It becomes garbage in garbage out because AI does give wrong information sometimes. So imagine you asking AI a question and that, you know, the response is false, because you are not smart enough, you are not even bothered, you know, you don’t even know what you’re supposed to do. So you don’t even know if that response is wrong, and you copy and paste, you are failed. So for the smart people, for the people who are ready to work, AI makes their work faster. I mean, it’s straightforward. So to that question, yes, you can actually use AI responsibly. And I’m certain you wouldn’t have problem with your lecturer. You want to speak to that?

Sade Dada: I do. I think, Martha, you’ve done an excellent job and professor as well. But just to add two points to this. The question, he started off by talking about like the professors that are specifically saying you can’t use AI. And I think we we need to do a little bit more around helping university professors understand what AI is and what it is. And where the world is is that there’s a lack of understanding even in the class of professors that we have that are making these judgments. Long before I joined META, I was an instructor at Indiana University in the US in their telecom department. And we were then struggling with people using the internet to come up with portions of their papers. And, you know, there’s a lot of hoopla around this. And that’s when you started seeing websites that you can go on that instructors, professors can go on and put in language that they received from students to kind of see where else this has shown up around the world. And it was a big deal then to be able to do that. Same thing where, you know, students were using all these online tools to give them summaries of books that they should be reading. And that was a big thing back then. But now no one really is talking about that. No one really cares about that because we’ve moved on. And there’s been an evolution in how we use new technologies and access to resources that we have. And so the second point I would like to make is that we we don’t want to be left behind in the world. Right. So when we’re when we think about the use of AI to help in educational studies, we don’t want a situation where the rest of the world is using it and they’re getting ahead because of that. Because, you know, in universities, research is a very big deal. And having publications is a really big deal. It helps you to go further in either your career or if you want to be a lifetime academic. And so we don’t want to find ourselves in our region in a place where others who are embracing AI get an advantage in the global stage. And, you know, as our countries are looking at how to use our digital skills and our digital knowledge as a. Um, we, we can’t, uh, miss that opportunity because people are scared of, uh, folks not learning because they’re utilizing the tools rather than gaining the concepts.

Martha Omoekpen Alade: Thank you very much, Sade. So, it’s obvious, we, AI has come to stay. There’s no shortcut to it. The, the issue is using it responsibly. Responsible use of AI, that’s where we should be focusing on, not whether we should use or not to use. We’ve passed that stage. So, um, this, the second question, I think it was asking about AI digital divide. So, um, Prof, could you, uh, respond to that, to the second question?

Wale Adedokun: Thank you. Uh, if I get that question very correctly, um, I’m sure some of the things that we, we have discussed answer that question. Um, digital divide basically has to do with what is happening, probably in the West, what is happening here. And like I said, if we want to, um, make use of AI, that speaks to our needs, we have to develop the system, the model that understands our needs and our temporary, uh, environment or situation, diseases, um, whatever in our environment. And the way to start really is to begin to, it’s not that we’re not developing models, we are developing models. If some of the models we develop, the West use them, but they are more relevant to them in the West because the data that was used to develop them came from the, from the West. So, Those same models developed here can be used here if we develop them or we train them and test them with local data. So the issue here really is data sets. Sincerely, the major issue is data sets, whether it has to do with funding. Funding is also for data sets. Whether it has to do with people in all of our data sets. The minute we begin to collect, curate, save, and make available localized data sets, I’m sure quite a number of these devices that we’re looking at will visit us regularly.

Martha Omoekpen Alade: Thank you very much, Prof. Prof, I want to ask a question. What you have said and what we have deliberated on now, do you think we could push this, that we need maybe funding for data governance training for researchers, maybe across Nigeria? Do you think that makes sense? Or what do you think?

Wale Adedokun: Let me have your thoughts on that. It makes a lot of sense, but you see, when you want to push for funding, there is a need to show what you do. There is a need to be able to prove a concept, basically, to be able to show what you want to use the funds to do, or what you have been able to do with the little that you have. So when you have funds, then you can expand beyond this.

Martha Omoekpen Alade: Let me interrupt you. What I’m asking is this, you as an expert, you have identified the problem, and we have also identified by the over series of panel sessions. We know one of the problem is, in fact, the major problem is the data set. And I tell you, we have data. We have data, but the data, they are not findable. They are not open. And a lot of people that have this data, they don’t know how to harness this data to bring it and make it available. Let me just break down the English. So if we say, okay, let’s form a committee and look into how we can get experts from across the country. You are an expert in that field. Experts from outside the country will bring experts together to see how we can organize, even if it’s an online training for researchers across Nigeria to train them on proper data governance. What do you think about that?

Wale Adedokun: I agree with you completely. Like I was saying, which you also alluded to, we have data in silos, different places, I mean, all over the place. Some of this data that we have, they are not yet in the proper formats, not where curated, because you have to go through the process for this data to be ready for anybody to use. You can use them raw, but for anyone to use them, they must be properly curated. So I agree with you maybe, and I think the Honorable Minister of Communication and Dictate Economy, I don’t quite a bit in putting together. I am aware some of my colleagues and some of the committees, putting them together to averse data from different sources. And eventually, they will not be able to ensure that this data are in. the proper formats that they are supposed to be they put them in public place for them to be used so with funding you know pulling people together also has to do with a lot of fun so with funding this can be extended to uh not only the kind of data we are looking at maybe in the education sector we begin to look at data from the hospitals data from uh the agricultural sector start from several other places not only technical people even medical medicals and even farmers local farmers if you train them on the kind of data you want they will collect this data for you somebody was talking to me about venom snake venom we have local people that connect that collects venom they call them is this snake or whatsoever they collect a lot of data and if you get these people you get all this data from them then you can prepare them in such a way that it could be useful to people so uh i agree with you that we can pull people together and see how they can harness this data and present it to whosoever is going to fund whether a body whether it is governments that will have this with funds we can pull them together we can go into places pull this data together and make them accessible to to people for research

Martha Omoekpen Alade: yeah i quite agree with you thank you very much prof especially in that area of open data and if these researchers are well um if they are trained and they understand this data governance it will be easier for them to they are the one collecting the data from the grassroots so whoever they want to collect the data from they will be able to in turn train those people to say this is what we want this is how we want you to collect the data right now my organization women technology in nigeria we are collecting data across the socio-political zones and i know what i’m going through I’m working with teachers, and they are collecting data. They’ve been submitting, and each time they submit error, error, so I have to keep training, retraining. It’s not easy. It’s not easy. It’s stressing me out. So, it’s a big deal, and I think we should look into it, and I hope that as one of the outputs of this program, it will be pushed beyond this gathering. So, is there anybody online that wants to say something? Any questions from online? No questions. Okay, so I think we are safe to close. We are just on time. So, thank you to our audience for your excellent questions. Please, a round of applause for our panelists. They did a good job for their invaluable contributions and insights today. So, Ms. Sade Dada, Professor Wale Adedokun, Mr. Salieu Kanu Mansaray couldn’t make it, and I don’t know why. Mr. Isola Olorunisomo couldn’t make it as well. Your expertise has truly enriched this plenary session, offering tangible pathways for AI, data governance, and innovation for development in West Africa. So, we’ve heard critical discussions on balancing innovation with safeguards, developing indigenous solutions, mitigating cyber security. The cyber security experts couldn’t make it, but we know that cyber security has been covered in previous sessions as well. And it’s imperative for regional harmonization. So, these discussions reinforce the immense potential and the critical responsibility that comes with leveraging AI for sustainable development in our region. The recommendations and the opportunities identified today will undoubtedly contribute to the future direction of policy and practice in West Africa. So on behalf of the West Africa Internet Governance 2025 Organizing Committee, thank you all for your participation. Thank you very much.

Agreement points

Importance of local data for AI development

Speakers

– Sade Dada
– Wale Adedokun

Arguments

Improving connectivity and access to enable AI adoption

Developing inclusive and culturally relevant AI datasets

Summary

Both speakers emphasized the critical need for local, culturally relevant data to develop AI solutions that address regional challenges.

Need for collaboration in AI development and regulation

Speakers

– Sade Dada
– Wale Adedokun

Arguments

Creating an enabling regulatory environment for AI innovation

Engage communities when collecting data and developing AI solutions

Summary

Speakers agreed on the importance of collaboration between different sectors and stakeholders in developing AI regulations and solutions.

Similar viewpoints

All speakers emphasized the need for responsible and context-specific AI development and use, particularly in addressing local challenges and in educational settings.

Speakers

– Sade Dada
– Wale Adedokun
– Martha Omoekpen Alade

Arguments

Improving connectivity and access to enable AI adoption

Develop problem-driven research focused on local challenges

Responsible use of AI in education

Unexpected consensus

Balancing AI use in education

Speakers

– Wale Adedokun
– Martha Omoekpen Alade
– Sade Dada

Arguments

Develop problem-driven research focused on local challenges

Responsible use of AI in education

Creating an enabling regulatory environment for AI innovation

Explanation

Despite initial concerns about AI use in education, there was unexpected consensus on the need to embrace AI tools while ensuring responsible use and maintaining academic integrity.

Overall assessment

Summary

The main areas of agreement included the importance of local data, collaboration in AI development and regulation, and responsible AI use in various sectors, particularly education.

Consensus level

There was a high level of consensus among the speakers on key issues, suggesting a shared vision for AI development in West Africa. This consensus implies a strong foundation for future policy-making and implementation of AI initiatives in the region.

Different viewpoints

Unexpected differences

Overall assessment

Summary

There were no significant disagreements among the speakers. The discussion focused on complementary aspects of AI development and adoption in West Africa.

Disagreement level

Low level of disagreement. The speakers generally agreed on the main issues and challenges, with slight differences in emphasis or approach. This suggests a shared understanding of the key priorities for AI development in the region, which could facilitate coordinated efforts to address these challenges.

Partial agreements

Similar viewpoints

All speakers emphasized the need for responsible and context-specific AI development and use, particularly in addressing local challenges and in educational settings.

Speakers

– Sade Dada
– Wale Adedokun
– Martha Omoekpen Alade

Arguments

Improving connectivity and access to enable AI adoption

Develop problem-driven research focused on local challenges

Responsible use of AI in education

Key takeaways

Creating an enabling regulatory environment for AI innovation requires collaboration between sectors, harmonization of existing regulations, and problem-driven research focused on local challenges.

Improving connectivity and access is crucial for equitable AI adoption, requiring innovative funding models, spectrum allocation policies, and addressing digital literacy and affordability.

Developing inclusive and culturally relevant AI datasets is essential, including creating local datasets in indigenous languages, establishing ethical data practices, and providing funding and infrastructure for data collection.

Responsible use of AI in education involves developing frameworks for appropriate use, training professors, and balancing AI assistance with original thinking.

Resolutions and action items

Develop a framework for responsible use of AI in tertiary institutions (mentioned as already in progress)

Consider organizing training on data governance for researchers across Nigeria

Unresolved issues

How to effectively collect and curate local datasets across various sectors

Specific funding mechanisms for AI research and data collection efforts

Detailed strategies for bridging the AI digital divide between West Africa and more developed regions

Concrete steps for harmonizing AI regulations across West African countries

Suggested compromises

Allow students to use AI for assignments, but require them to submit the prompts used to demonstrate understanding

Balance the use of AI in education with the need for original thinking and work

When it comes to an enabling regulatory environment, what I’ve seen in the course of my work that is very useful is when we focus on co-creation. And what do I mean by that? That means that you collaborate with various parties that impact a particular regulatory agenda or a particular field or innovation sector.

Speaker

Sade Dada

Reason

This comment introduces the important concept of co-creation in developing AI regulations, emphasizing the need for collaboration across sectors.

Impact

It shifted the discussion towards a more inclusive approach to AI governance, highlighting the importance of involving multiple stakeholders in the process.

Currently now, as we speak, some of our undergraduate students are in China, representing Africa, with a product that they just developed, and the product is called Crop Disease Detector.

Speaker

Wale Adedokun

Reason

This example demonstrates a practical application of AI developed by African students to address local agricultural challenges.

Impact

It grounded the discussion in real-world examples and showcased the potential for African-developed AI solutions, inspiring further conversation about local innovation.

We need to begin to look at, if we want our researchers to be very relevant, start to develop local datasets. So let’s look at datasets, whether it is in agriculture, it’s in education, it’s in health. And not just create these datasets in English language, we also need to begin to see how we can create datasets in our local languages.

Speaker

Wale Adedokun

Reason

This comment highlights the critical need for locally relevant and linguistically diverse datasets in AI development for Africa.

Impact

It deepened the conversation about data collection and curation, emphasizing the importance of cultural and linguistic relevance in AI development for the region.

We don’t want to find ourselves in our region in a place where others who are embracing AI get an advantage in the global stage. And, you know, as our countries are looking at how to use our digital skills and our digital knowledge as a. Um, we, we can’t, uh, miss that opportunity because people are scared of, uh, folks not learning because they’re utilizing the tools rather than gaining the concepts.

Speaker

Sade Dada

Reason

This comment addresses the potential consequences of not embracing AI in education and research, framing it as a global competitiveness issue.

Impact

It shifted the conversation towards a more strategic view of AI adoption, emphasizing the need to balance concerns about AI use with the imperative to remain competitive globally.

Overall assessment

These key comments shaped the discussion by emphasizing the importance of collaborative approaches to AI governance, showcasing practical examples of African AI innovation, highlighting the need for locally relevant datasets, and framing AI adoption as crucial for global competitiveness. The conversation evolved from theoretical discussions to practical considerations and strategic imperatives, providing a comprehensive view of the challenges and opportunities in AI development and adoption in West Africa.

How can we develop more inclusive, culturally relevant datasets for AI in Africa?

Speaker

Martha Omoekpen Alade

Explanation

This is crucial for addressing AI bias and creating AI solutions that are relevant to African contexts and problems.

What infrastructure is needed to support the development of local datasets in Africa?

Speaker

Martha Omoekpen Alade

Explanation

Infrastructure is essential for collecting, storing, and sharing local data to train AI models that are relevant to African needs.

How can we create a central repository for localized data in Africa?

Speaker

Wale Adedokun

Explanation

A centralized data repository would facilitate sharing of local datasets among African researchers and AI developers.

How can we develop datasets in local African languages?

Speaker

Wale Adedokun

Explanation

This is important for creating AI models that can understand and communicate in local languages, making them more accessible to diverse African populations.

How can we improve data governance training for researchers across Nigeria?

Speaker

Martha Omoekpen Alade

Explanation

Better training in data governance would help researchers collect, manage, and share data more effectively, supporting the development of local AI solutions.

How can we balance aspirational AI development with solving fundamental issues in Africa?

Speaker

Jose from Ghana (audience member)

Explanation

This addresses the need to develop advanced AI capabilities while also addressing basic infrastructure and development challenges in the region.