Kashifu inuwa is the Director-General of Nigeria’s National Information Technology Development Agency, NITDA, believes that Africa’s best bet for winning the global AI race is not to catch up on massive infrastructure but rather to build talent and develop small language models (SLMs), tailored to local reality. According to him, Africa has a unique opportunity to leapfrog infrastructure barriers that have historically prevented it from participating in global tech revolutions.
Nigeria, despite its rapid digital adoption, has not been a top priority for recent AI Infrastructure investments, such as those made by Cassava Technologies and Nvidia, or MediaTek in Egypt. Inuwa explains that this is not due to a lack ambition, but rather to timing and strategic preparedness. He cited Nigeria’s Local Content Guidelines from 2016, revised in 2018 and 2019 which laid the groundwork early by encouraging data residence. Since then, NITDA launched an eight-pillar road map focused on talent development, research, infrastructure and regulation, innovation, and regulation, all geared towards fostering a vibrant, digital economy.
Inuwa says that instead of trying to compete with hyperscalers such as OpenAI and Google, Africa should focus on AI models and applications that solve local issues–from healthcare to agriculture.
He says that “we don’t need a huge infrastructure to build models,” citing the development by EqualyzAI, a Nigerian startup, of Afro-SLM, which is a lightweight AI engineered specifically for African languages, offline settings, and rural communities. Inuwa says that such models could democratise representation and access to AI systems.
NITDA, in partnership with Google is laying the foundation for Nigeria to become an AI and cloud hub. Discussions are taking place around a sovereign infrastructure for cloud computing, investment frameworks and a pipeline of talent in Nigeria to support it. Inuwa sees Nigeria as a hub for West and Central Africa that will not only digitise government services but also attract hyperscalers and enable semiconductor manufacturing. Inuwa, however, says that Nigeria must focus not only on infrastructure, but also on ecosystems to get there. This means aligning data classification policies, investing in AI across sectors, and integrating the digital skills from kindergarten onwards. According to him, the AI of the future is about local intelligence and strategic partnerships. It’s also about inclusive innovation, which makes African data count when it comes to global decisions. He says, “We missed the first three industrial Revolutions.” “We cannot afford the fourth.”
TechCabal interviewed Kashifu inuwa for more information on the details of their discussion with Google and the push to digitalise government services. This interview has been edited to make it shorter and clearer.
Cassava Technologies, in collaboration with Nvidia, has made a couple of digital infrastructure investment into Africa. MediaTek funds an Egyptian startup in order to grow the AI semiconductor market on the continent. Why is Nigerian not a top priority for these investments given the size of the country in digital adoption?
Let’s go back in time. If you ask my personal opinion, I believe that Nigeria made a conscious decision to position itself as a hub for IT infrastructure across Africa. Local Content Guidelineswas first drafted in 2016 and was revised in 2018 and 2019. The Local Content Guidelines encouraged agencies to keep their data in-country.
We developed our strategic roadmap and plan at NITDA based on this vision, with the vision of making Nigeria an empowered nation through digital innovation. We came up with eight pillars. The first is to cultivate digital literacy and talent, since it’s a know-how-based economy. The human component of technology is talent. People make technology better.
Hyperscalers need talent to build their services if they are coming to Nigeria. We are doing a great deal based on that, such as working with the Minister for Education. We have developed a program to integrate digital literacy into formal education, from kindergarten to tertiary schools. We are implementing an acceleration program for Nigerians to learn technology skills that are in high-demand.
Building a robust research eco-system is the second pillar, because literacy and talent can be quickly achieved. We need to build this deep tech ecosystem and research for the future because it will help the nation build a robust, sustainable economy. We are focusing on AI, Internet of Things, Unmanned Aerial Vehicles, Blockchain, Robotics, and Additive Manufacturing. The third pillar focuses on creating a legal environment that fosters innovation. The fourth pillar is to promote inclusive access to digital services and infrastructure, because the more digitally-literate we are, the better it will be for the digital economy. We also have a pillar for cybersecurity and digital trust. We have a pillar on entrepreneurship, an ecosystem of innovation and entrepreneurship. We have a strategic partnership pillar and a digitalisation of government pillar.
Google liked the idea (strategic road map), and we agreed on five pillars that we could work on together, including scaling infrastructures, cloud adoptions, AI innovation and investment framework. If NITDA, Google and other hyper-scalers get this right, they will come to Nigeria and invest. It will allow our local data centres to grow. It will also open doors for other technology companies like Nvidia to come to Nigeria because it’s all about creating an ecosystem.
We are not only focused on technology, but also horizontal value creation in all areas. NITDA is examining how we can create sustainable energies because it’s an area we need to pay more attention to if want to attract hyperscalers. It’s not just about the grid when it comes to energy. We must also be innovative, because even developed nations are building data centers that use off-grid power sources.
Next, we have the AI integrations because AI will be the next big wave in our history. It will disrupt and change many things. AI is all about data, so we won’t just use AI developed by other nations. If your data is skewed then so will your output. Data does not reflect your values or culture.
AI will not take you into account when making decisions. We need to ensure that every citizen is visible digitally in the AI systems that we are building. Imagine building a system to decide who gets what within a society. It takes more than a software engineer sitting in a cubicle. You need data. We need to connect people. We need to collect pain data. We need to use data to train AI models or retrain them.
To achieve what we’re trying to do, we need eight hyperscaler data centres. We also want to position Nigeria, as a hub in West and Central Africa because of the landlocked countries that surround us, to be able to achieve this sovereign cloud infrastructure. How can we leverage our opportunity, the submarine cables, and the 90,000 kilometers of fibre we will deploy, to extend services to neighboring countries?
What can we do to position ourselves as the leader? NITDA is considering a data classification law or policy. If I say that my data won’t be sent outside of the country, then how can I expect my neighbor to trust me enough to give me his data? You can see the paradox, right? We need to balance this. We need data classification to know which data is allowed to stay in the country and which can be sent to the public cloud. If other nations bring their data, they can also bring Nigerian data into our cloud infrastructure.
This will open up many things. We saw what happened in the banking sector two years ago, and more recently. Most banks are required by law to keep their data in the country. Some banks have data centers and others use local data center providers. However, they lack hyperscale capability. We saw what happened when the banks tried to implement the cashless policies. They were stressed. Some of them have recently upgraded, and they are still struggling to get back to where they used to be six months ago. They have a local cloud infrastructure in the country. I don’t believe there is a law that prevents them from going cloud. Imagine that our oil and gas industry, as well as government, also went cloud because the president gave a mandate to digise government services.
The cashless policy of the government has made Nigeria a fintech center. When the central Bank of Nigeria pushed for banks to build technology infrastructures and to digitalise, gaps emerged, and technology companies jumped into action and began building solutions. Imagine the opportunities that will be created if we digitalise all government services. This can only be done if we have local cloud capabilities. Something will happen soon with the conversation with Google.
What percentage of government data is stored in the cloud? What efforts have been made to ensure that government agencies comply with digitisation regulations?
According to our assumption and extrapolation, only 5-10% of government services are digitalised. Many services aren’t. Galaxy Backbone is responsible for government data services, but NITDA recently issued Digital Public Infrastructure Standards (DPI), because we believe this will lay the foundations for the digitalisation of government service. This is the same approach that some countries have taken recently, such as India.
According to the DPI standards you need four components to easily digitalise government services. First, you will need a digital ID that can be scaled to become the foundational ID of our digital public infrastructure. You will also need a payment infrastructure. NITDA collaborates with the Nigerian Interbank Settlement System, which is also very strong when it comes to providing payment infrastructure. It allows fintechs and startups to create solutions and connect with each other.
You need a platform for data exchange. In Nigeria, we lack this platform. We have data stored in silos because of this. The standard will provide the framework for creating this platform. We have already started building it internally. In the financial sector we are working with Nigerian Inter-Bank Settlement System to implement ISO standards on data, etc. We are also working together with the Nigerian Financial Intelligence Unit to implement a standard that will improve visibility of financial transactions.
NITDA also tries to leverage Global Public Benefit (GPB infrastructure) to create a platform that will allow government, ministries and departments to share data. This will be the foundation of DPI. To be able to use a service, you will need to first identify yourself. Second, all services, whether they are public or private, usually come with a charge. You have to pay for the service you use. You need to be able exchange data seamlessly between MDAs, Ministries, Departments, and Agencies. You don’t have to build large systems with the way things are evolving today. You can use AI for building agents that will help you fetch data and present it in the way you desire.
What role does Nigeria play in the global AI war led by the United States of America and China? Is the government considering being involved in the AI manufacturing process?
We need to be smart. Africa, not only Nigeria, suffered during the first, third, and fourth Industrial Revolutions. Why? To be a part of it you need to make a large initial capital investment. The fourth Industrial Revolution is different from the previous three. It’s a revolutionary change because it’s knowledge-based. Even if you have technology, you still need talent to improve it.
I know that people say AI will replace humans or do this and that. I understand that AI will be a challenge. However, I don’t believe AI will replace humans like you and I. It will replace our skills and the processes that we use to do our jobs. We must therefore reskill. We need to learn to use AI to be more relevant. We need a human to be in the loop, as it is. You can’t just leave everything up to AI.
We need to be aware of what is happening in Africa. We shouldn’t rush in, because it could be like Alice in Wonderland. They are currently experimenting. They don’t yet know what will happen or what they need. OpenAI and co. have invested heavily in infrastructure for LLM. DeepSeek popped up and said that you didn’t need all this infrastructure to build LLM. You can distill and get it t o perform as well as those with large infrastructure.
What will happen if, with our limited resources, we rush in with these large GPUs and then realize that we don’t require as much to build our LLM?
What is the best way to democratise access to education?
By making it cheaper. How can you make something cheaper? By lowering the prices and coming up with innovative ways to get more for less. I think we should be focusing more on developing talent and also application while watching the infrastructure race. But we must also be mindful of building the ecosystem.
The goal is to create an ecosystem and not just have someone come in and build a datacenter for us. In Nigeria, you can find most of the raw material needed to make these semiconductors.
I think Africa should be focusing on the idea of a small language model. How can we achieve the best results with less infrastructure? In terms of application, however, I believe that we can lead in Africa because it will allow us to leapfrog. We’ve done it with mobile technology. I believe we can do it with AI.
Here at NITDA we have developed a roadmap for AI transformation. We have already started implementing the roadmap. We want to leapfrog. We don’t have to follow the digital maturity ladder. We are currently at level 2, but we won’t be able to jump to AI (level five) because we still have a lot of legacy systems in place. We don’t have to follow them. We are also looking at the bigger picture, not only adopting technology but also the evolution of jobs. We’ve done amazing work in this area without realizing it.
For example, in the government we have registries that receive physical documents and letters. We simply stack them into a book. We open the file and move it from one office into another individually. In the last three years, NITDA has automated its processes. We taught our staff how to scan documents. They scan the document and upload it to a system. We have an AI-based system where they can upload. The system will automatically email the document after reading the content. It has developed this roadmap without them or us, the management, even knowing. We now have to train our staff. People must now find a way to transition or evolve their job.
Are you close to implementing the work you are doing with Google?
NITDA’s work with Google has a timetable. We hope to have a clear implementation plan before October. NITDA has developed the vision for partnership. We’ve developed it. We met with the Nigerian president and the Google CEO to discuss the partnership and shared vision. We’ve already done the technical work. We are now about to build an investment framework, where we will bring in our local data centers, our private equity, and our VC. Google can also bring in investors.
We are willing to accept investment from financial development institutions. This framework will guide us on when we can begin the implementation. This framework should be complete, and we will know what to do next. We have already done the sizing and identified the demand elements. This is a private project we are doing together with Google. They don’t want it made public until they are certain of the business possibilities.
Once that is ready, anyone can be a part of it. We also want everyone to be able to participate. We want local data center providers involved. We are looking into how the government could bring in a sovereign investment. This work will create the eco-system. We then move on to the supply chain.
This is more than just hyper-scalers. We need to include even the technology equipment producers. We need to think about how we can manufacture some components of the technologies locally.
Could you briefly describe the research centers you said would be established?
NITDA works with universities to research Deep Tech. Not just research which will end up on a shelf collecting dust, but rather research that can be put into practice. In November, the International Conference on Electronic Governance will be held in our city. It’s an United Nations program that focuses on Nigerian digital infrastructure. We want Nigerian and foreign researchers to work on how Nigeria can create a sustainable digital public infrastructure.
NITDA wants to also research the application. How can we boost Nigeria’s agri-productivity using AI? How can AI be used in finance? How can we use blockchains, cryptocurrencys, etc.? Most of the six technologies I have mentioned are general-purpose. They are applicable to all critical sectors. It is important to do the research one at a time.
What can we do to improve healthcare using technology? How can we use AI in remote locations to deliver quality healthcare services? Imagine training an AI bot with data from Nigerian hospital consultations and doctors’ consultations. You can have a nurse attend to patients and then use the system to prescribe. Even if you use ChatGPT and compare the results with what a doctor will tell you, there is a very small difference.
