Published on 15 Feb 2022

Africa wakes up to the potential of artificial intelligence

How Singapore can stake a claim and take advantage of untapped opportunities

By Rafiq Raji


Ask an average African about artificial intelligence (AI) and the image that would likely immediately come to mind would be the futuristic and technologically advanced Wakanda  - the fictionalised utopia depicted in the 2018 blockbuster Black Panther. Ask about Apple’s digital assistant Siri, taxi-hailing app Uber, or even the deceptively simple Google search and you would almost certainly get a knowing nod and probably a reel of facts about technologies that touch the daily lives of ordinary Africans. These technologies are underpinned by artificial intelligence - the ability of a digital device to perform tasks commonly associated with intelligent beings.”[1] They are AI, combining “large volumes of data with computing power to simulate human cognitive abilities such as reasoning, language, perception, vision and spatial processing.”[2]

Numerous taxonomies exist for AI. One classifies AI as (1) Basic AI or Artificial Narrow Intelligence (ANI), (2) Advanced AI or Artificial General Intelligence (AGI) and (3) Autonomous AI or Artificial Super Intelligence (ASI) (Strusani & Houngbonon, 2020). Most of the AI in commercial use today like speech or facial recognition are ANI or Narrow AI, which are essentially ‘intelligent’ systems designed for doing one particular thing.”[3] AGI and ASI, which are AI with near-human and ‘super-human’ intelligence capabilities respectively, are still aspirational.

AI is not just one thing or the other. It is an enabler of many industries and facets of human life: scientific research, education, manufacturing, logistics, transportation, defense, law enforcement, politics, advertising, art, culture, and more (Kissinger, Schmidt & Huttenlocher, 2021).[4] Even so, fears about AI’s potential for causing Frankenstein-like harm to mankind - long built into the imagination of many through media, films and literature – is not outside the realm of possibility. Left unchecked, an advanced AI system could develop ‘logic’ that could one day act in conflict with human values and institutions. Unrestrained AI could put us all at peril as much as it make our lives easier.

In November 2021, all 193 member countries of the United Nations Educational, Scientific and Cultural Organization (UNESCO) adopted the first-ever global agreement on AI ethics to ensure that the technology evolves in a manner that is beneficial to humanity.[5] Africa certainly stands to gain from such global governance mechanisms to guard against negative outcomes of AI.[6] Even so, Africa is under-represented in AI – unlike Singapore, which is a leading AI innovation hub in Asia. Scarce AI talent, weak institutions and poor infrastructure are some of the reasons why Africa trails in this emerging technology. In this paper, I argue that Africa could leverage on Singapore’s AI expertise to address some of the constraints.

Africa is underrepresented in AI

A growing concern about AI is its ‘lack of diversity’, what Mohamed, Isaac and Png (2021) call “algorithmic coloniality” in Massachusetts Institute of Technology (MIT) academic Daron Acemoglu’s 2021 book Redesigning AI. In this reality, the logics, institutions, and practices of AI development and deployment are characterised by ‘oppression, exploitation and dispossession’.[7] Mohamed et al.’s (2021) ‘de-colonial AI’ would not interpret pictures of Africans as apes, Asians as having their eyes closed or concentration camps as ‘sport’ or ‘jungle gym’, says Oxford University economist Daniel Susskind (2021).[8]

Reducing AI bias would require ensuring that datasets that underpin its applications are wide-ranging and inclusive. AI trained with African faces as well as non-African ones is not likely to make racist interpretations, nor would it mistake facial features for facial actions, or mass forced detention as a ‘life-enhancing’ gathering. With China only second to America in the development of AI, the risk of algorithmic dispossession during the AI research and development phases for Asians is limited. AI systems and applications that rely on Asia-rich data are not likely to be easily amenable to algorithmic oppression and exploitation of Asians. Africa is yet to be similarly empowered.

Global AI expertise is dominated by the West and Asia. In fact, China is increasingly acknowledged as the global leader in AI, even ahead of the United States, some argue.[11] Africa’s representation is minuscle (see Table 1). There is also evidence of deliberate exclusion of Africans from the global AI innovation ecosystem. Speech recognition systems from five of the world’s biggest tech companies — Amazon, Apple, Google, IBM and Microsoft — make far fewer errors with users who are white than with users who are black.[12] African AI researchers repeatedly face holdups when applying for visas to attend global AI conferences, for instance.[13] This should not be the case for a technology that is soon to dominate our lives. New and emergent technologies tend to have a magnified effect on Africans due to the continent’s relatively lower developmental base. Unlike AI, most of these innovations are data-agnostic. A smartphone would work anywhere it can get a signal. A computer software can be downloaded anywhere from the cloud. The internet is available to anyone with a data connection. For AI to work well for Africans, however, it must use African data, have African characteristics (Apple’s Siri could sound African or be voiced by an African, say), and have African operators for the many implementation shortcomings that still exist with the technology.[14] One example is Common Voice, a free and public database of crowd-sourced voices of African languages for use in AI applications.[15]

The fact of the matter is that AI is not a priority for most African governments. But a few viz. Mauritius, South Africa, Seychelles, Rwanda, and Senegal are trying to establish their stakes in this emerging technology. South Africa issued a world-first patent to an AI system called DABUS (“Device for the Autonomous Bootstrapping of Unified Sentience”) in July 2021, for instance.[16] In November 2018, Mauritius published an AI strategy document.[17] To boost its key investment and asset management industry, the Mauritius Financial Services Commission issued rules for robotic and artificial intelligence enabled advisory services in June 2021.[18] Rwanda, which established a centre of excellence in the areas of digitalization and AI in November 2020,[19] is already leveraging on the technology for disaster management.[20]

Senegal’s Diamniadio smart city being built close to capital Dakar will rely on AI to manage public infrastructure and services.[21] In June 2021, the Senegalese government set up a Digital Technology Park in Diamniadio, where all of the state’s digital data and platforms is to be located. The facility was built by Chinese tech-giant Huawei but is to be managed by the State Informatics Agency of Senegal.[22] IBM, which has AI research labs in Nairobi, Kenya (opened in 2012) and Johannesburg, South Africa (added in 2015), is working on creating an African AI ‘ecosystem’ that is able to pull its weight globally.[23] In 2019, Google opened its first African AI lab in the Ghanian capital Accra. The tech giant also supports a graduate programme in machine intelligence at Rwanda’s African Institute for Mathematical Sciences (AIMS).[24] In September 2021, Andela, which connects African tech talent with global firms, announced a US$200m investment by Japan’s SoftBank that would allow it to enhance its services using AI.[25], [26]

Scarce talent, weak institutions and poor infrastructure

It would be fair to say that AI expertise in Africa is rather limited, with few practitioners acquiring skills through formal education. The little expertise that does exist is mostly self-taught or comes through work experience.[27] The International Finance Corporation (IFC) estimates the actual number of tech talent across the continent to be 690,000 (2019) – mostly available in South Africa, Egypt, Nigeria, Kenya, and Morocco (see Table 2).[28] UNESCO, which published the findings of its AI needs assessment survey for Africa in April 2021, found that AI governance policies, legal and regulatory frameworks, and talent capacity in Africa needs to be strengthened.  The UN body says AI priorities should be ‘harmonised’ at a ‘continental level’, and recommends that greater effort should be put towards ‘advancing AI education, research and training’.[29] There are already initiatives afoot to bridge some of these gaps. In September 2021, for instance, South Africa’s University of the Witwatersrand took the lead in establishing the AI Africa Consortium - a partnership between academia and industry to build AI research and application capacity on the continent.[30] This addresses a much-repeated refrain about limited synergy between the private and public sectors on the one hand, and academic and industry on the other.[31], [32] 

While basic AI can be applied using 2G mobile technology, which is widely available across Africa, the exponential value-add from AI can only be realised with 5G technology (and the advanced ones that are likely to follow in the future). Some African countries have already started building the foundations for 5G services.[33] But basic hard infrastructure like reliable power supply and robust telecommunications would have to be made available more widely if Africa is to successfully unleash the potential of AI (Figure 2 and Figure 3).

Progress is being made. AI-based smart electricity metering has been introduced by many state-run utilities as a way to tackle illegal consumption, inaccurate billing and poor customer service.[34] A global drive towards renewable energy also increases the urgency for AI-powered smart grids on the continent, which are better suited to manage an energy source that is both intermittent and distributed.[35] Global consulting firm BCG believes telecommunication firms would be better placed to manage “fluctuating demand levels, adjust to supply chain disruptions, and adapt to sharp shifts in consumer confidence and priorities” using AI.[36] Some of these challenges are magnified in many African markets, where mobile telecommunication towers are sometimes powered entirely by standby generators, secured by armed guards, and are not easily accessible by road. Thus, while AI would only thrive when the requisite power and telecommunication infrastructure is in place, AI would also increasingly be needed to build and maintain this infrastructure.

The prospects of AI for the improvement of education and learning outcomes in Africa are significant.[38] AI enables teacher productivity as it enables them to reach more students via online platforms that are easily customised for each student, as routine tasks like grading and providing feedback are automated, allowing them more time to do those tasks requiring human input exclusively. African healthcare professionals are increasingly able to do more with less owing to AI as well, a development writ large during the early challenging months of the Covid-19 pandemic.[39] Diagnosis can be done remotely on AI platforms, saving doctors time and allowing them to focus on complex cases for which physical consults are necessary, for which they are also able to collaborate remotely with leading global consultants and surgeons in real time when cases warrant it.

AI is also helping banks assess credit-worthiness of their customers – particularly small depositors and small businesses – much more efficiently.[40] FinTech firms, for instance, are able to do reasonably accurate credit scoring by feeding alternative data points into AI, allowing banks to provide credit to the once financially excluded. AI is also transforming smallholder agriculture by providing real-time data on soil conditions, seeds, pesticides, and spot prices.[41] AI could also help make Africa’s transportation systems safer, efficient and less chaotic.[42] AI in African manufacturing is controversial, however, as workers fear losing jobs to automation, even as it promises to make it far more productive.[43] But at the end of the day AI will fail to make impact in Africa unless the continent can nuture more AI talent, fix their legacy infrastructure problems, and strengthen their institutions.

Singapore-Africa AI partnership

Singapore, which is promoting the responsible development of AI, and was the first Asian country to establish a governance framework for AI, is an exemplar and a more appropriate archetype for Africa. [44], [45] Singapore was ranked the smartest city in the world in the IMD SCO Smart City Observatory Smart City Index 2021, and has ranked highest since 2019.[46] 'Smart cities’ leverage on digital technologies to improve the quality of life of their dwellers.[47] They rely on AI and other digital technologies to maintain and upgrade infrastructure and provide public services to improve performance and reduce costs.[48], [49] Thankfully, a budding partnership between Singapore and Africa is beginning to emerge over AI.

Take for instance Singapore-based Terra AI. It is working with Senegalese tech startup incubator CTIC Dakar and Kenya’s Cliniq to provide AI training programmes.[50] Singaporean blockchain technology firm InfoCorp is working with Kenya’s Lofte Kesho to tokenise livestock and thus facilitate financial inclusion.[51] When Covid-19 struck and the South African government needed a tech solution for contact tracing, it turned to Singapore-based AI firm Sqreem Technologies to help build a Covid-19 tracking and tracing platform.[52] Sqreem is now adapting AI for the financial services, defence and auto industries.[53]

Using mobile metadata, Singapore’s Credolab uses AI to provide alternative credit scoring services to financial institutions in a number of African markets such as South Africa, Nigeria, Kenya and Ghana, where lenders hitherto struggled to get credit scores for as much as 70% of their customers.[54] In August 2021, Singapore’s leading AI and big data firm Crayon Data announced it was expanding into South Africa by partnering with a local tech firm Falcorp Technologies that is helping banks provide more personalised customer experiences through the use of AI.[55]

But all of this is scratching the surface. There exists far more room for collaboration and partnerships between Singaporean and African AI firms. As in the case of Sqreem, applications developed in one African market - accidental or otherwise - could easily be adapted for others. African governments see in Singapore a model to create smart cities and a gold standard when it comes to AI policy formulation and governance.    


Africa’s underrepresentation in global AI provides an opportunity for Asian firms to tap business. Thus far, China has taken the lead but technologically advanced and exceptionally experimentative Singapore is not that far behind. Its AI governance standards, thought leadership, and tech know-how can help Africa bridge the gap. There is plenty of untapped business opportunities for Singaporean firms in Africa in this space. All that is required is the willingness to explore, invest and take risk.



[1] Copeland, B.J. (2021, November 9). Artificial Intelligence. Britannica. Retrieved from

[2] Strusani, D. & Houngbonon, G.V. (2020). The role of artificial intelligence in supporting development in emerging markets. In International Finance Corporation (Eds.), Artificial intelligence in emerging markets: Opportunities, trends and emerging business models. Washington DC: IFC. Retrieved from

[3] Manning, C. (2020). Artificial intelligence definitions. Stanford: Stanford University Human-Centred Artificial Intelligence. Retrieved from

[4] Kissinger, H.A., Schmidt, E. and Huttenlocher, D. (2021). The age of A.I. and our human future. London: John Murray.

[5] United Nations Educational, Scientific and Cultural Organization (2021, November 25). UNESCO member states adopt the first ever global agreement on the ethics of artificial intelligence [Press release]. Retrieved from

[6] Besaw, C. and Filitz, J. (2019). Artificial intelligence in Africa is a double-edged sword. Our World. Retrieved from

[7] Mohammed, S., Isaac, W.S., and Png, M. (2021). Decolonizing AI. In Acemoglu, D. (Eds.), Redesigning AI: Work, democracy, and justice in the age of automation. Boston: Boston Review Forum.

[8] Susskind, D. (2021). A world with less work. In Acemoglu, D. (Eds.), Redesigning AI: Work, democracy, and justice in the age of automation. Boston: Boston Review Forum

[9] Stanford University Human-Centred Artificial Intelligence (2021). Artificial Intelligence Index Report 2021. Stanford: Stanford University. Retrieved from 

[10] Bradshwa, N. (2020, October 13). Artificial intelligence in Africa – Will 2020 be a tipping point for African AI? AI Expo Africa. Retrieved from

[11] China has won AI battle with U.S., Pentagon’s ex-software chief says (2021, October 11). Reuters. Retrieved from

[13] Knight, W. (2019, November 8). African AI experts get excluded from a conference – again. Wired. Retrieved from

[14] Cisse, M. (2018, October 23). Look to Africa to advance artificial intelligence. Nature. Retrieved from

[15] Mureithi, C. (2021, May 30). Siri and Alexa still don’t support African languages. Quartz. Retrieved from 

[16] Naidoo, M. (2021, August 9). In a world first, South Africa grants a patent to an artificial intelligence system. Quartz Africa. Retrieved from

[17] Government of Mauritius (2018). Mauritius Artificial Intelligence Strategy: A report by the working group on Artificial Intelligence. Port Louis: Government of Mauritius. Retrieved from

[18] Mauritius Financial Services Commission (2021, June 18). The FSC issues the Financial Services (Robotic and Artificial Intelligence Enabled Advisory Services) Rules 2021 [Press release]. Retrieved from

[19] Rwanda Development Board (2020, November 25). The Government of Rwanda announces the establishment of a center of excellence in the areas of digitalization and AI [Press release]. Retrieved from

[20] United Nations Educational, Scientific and Cultural Organization (2021, November 12). Stakeholder training on AI to boost disaster management in Rwanda [Press release]. Retrieved from

[21] Djamessi, F. (2019, October 23). AI offers answers to Africa’s rapid urbanization. SciDev.Net. Retrieved from 

[22] Olander, E. (2021, June 24). The powerful symbolism of the Huawei-built data center deal in Senegal. The China Africa Project. Retrieved from

[23] IBM (2021, July 28). What’s next in AI: IBM Research Africa to host virtual AI seminar series in August [Press release]. Retrieved from

[24] Adeoye, A. (2019, April 16). Google has opened its first Africa artificial intelligence lab in Ghana. CNN. Retrieved from

[25] Andela (2021, September 29). Andela announces $200M investment led by SoftBank [Press release]. Retrieved from

[26] Ayugi, R. (2021, October 11). Nigeria startup Andela is now West Africa’s 3rd unicorn. Tech In Africa. Retrieved from

[27] Artificial intelligence for Development in Africa (2021). Artificial intelligence in sub-Saharan Africa: Compendium Report. Ontario: International Development Research Centre. Retrieved from 

[28] Google & International Finance Corporation (2020). E-Conomy Africa 2020: Africa’s $180 billion internet economy future. Mountain View & Washington DC: Google & IFC. Retrieved from

[29] United Nations Educational, Scientific and Cultural Organization (2021, April 2). UNESCO launches the findings of the artificial intelligence needs assessment survey in Africa [Press release]. Retrieved from

[30] Wits leads ambitious partnership to drive AI in Africa (2021, September 8). Retrieved from

[31] Candelon, F., El Bedraoui, H., and Maher, H. (2021). Developing an Artificial Intelligence for Africa strategy. Development Matters. Retrieved from

[32] Travaly, Y. & Muvunyi, K. (2020, January 13). The future is intelligent: Harnessing the potential of artificial intelligence in Africa. Africa in Focus. Retrieved from

[33] GSM Association (2019). 5G in Sub-Saharan Africa: laying the foundations. London: GSMA. Retrieved from

[34] Maime, R.B. & Mosala, T. (2021). Why smart meter adoption in Africa lacks cohesion. ESI Africa, 1, pp. 60-61. Retrieved from

[35] Makala, B. and Bakovic, T. (2020). Artificial intelligence in the power sector. EM Compass. Retrieved from

[36] Banerjee, S., Candelon, F., Lorenzo, R., Cacouros, A. and Harguil, S. (2020). Transforming telcos with artificial intelligence. Boston: Boston Consulting Group. Retrieved from   

[37] Lakmeeharan, K., Manji, Q., Nyairo, R. & Poeltner, H. (2020, March 6). Solving Africa’s infrastructure paradox [Blog post]. Retrieved from

[38] Agajo, J., Shuaibu, A.S. & Guda, B. (2020, December 16). Advancing education and speech recognition in Nigeria with AI and machine learning [Blog post]. Retrieved from 

[39] Akogo, D. (2021, March 22). Five ways AI can democratise African healthcare. Financial Times. Retrieved from

[40] Nieburg, O. (2021, July 20). Will AI risk analysis really expand access to credit in Africa? The Africa Report. Retrieved from

[41] Ordu, A.U., Cooley, L. & Goh, L. (2021, August 16). Digital technology and African smallholder agriculture: Implications for public policy [Blog post]. Retrieved from

[42] Conde, M.L. & Twinn, I. (2019). How artificial intelligence is making transport safer, cleaner, more reliable and efficient in emerging markets. EM Compass. Retrieved from 

[43] Chudasama, S. (2021, August 5). Leveraging AI and big data to address manufacturing challenges. CIO Africa. Retrieved from

[44] Granzen, A. (2021, February 5). Singapore’s model AI governance framework sets out to help organizations deploy AI responsibly [Blog post]. Retrieved from

[45] Sagar, M. (2019, January 25). Singapore releases first artificial intelligence (AI) governance framework in Asia. Open Gov Asia. Retrieved from

[46] International Institute for Management Development & Singapore University of Technology and Design (2021). Smart City Index 2021. Lausanne & Singapore: IMD & SUTD. Retrieved from 

[47] Wade, M.R. & Pfaffli, M.P. (2016, May). What is a smart city anyways? IMD. Retrieved from

[48] Brown, A. (2021, August 4). Developing coherent AI infrastructure for smart cities. Forbes. Retrieved from

[49] Echendu, A.J. (2021, March 8). Smart city technology: a potential solution to Africa’s growing population and rapid urbanization? Development Studies Research, 8(1). Retrieved from

[50] Ang, J. (2021, August 23). Singapore firm ventures to make AI accessible in Africa. The Straits Times. Retrieved from

[51] Enterprise Singapore (2021, August 23). Singapore companies’ projects in Africa doubled in five years [Press release]. Retrieved from

[52] Yuit, C.K. (2021, September 13). ‘Huge opportunities’ for Singapore businesses in Africa: Experts. Channel News Asia. Retrieved from

[53] A.I. shop goes from tracing Covid-19 to virtual banking (2020, April 6). Digital Finance. Retrieved from

[54] Nieburg, O. (2021, July 20). Will AI risk analysis really expand access to credit in Africa? The Africa Report. Retrieved from

[55] Crayon Data (2021, August 23). Crayon Data brings AI-led personalization platform,, to South Africa [Press release]. Retrieved from  

Subscribe to Newsletter