Artificial intelligence technologies and their challenges in Africa


Daniel Makina, University of South Africa

The focus of this note is an attempt to answer the question: What are the potential benefits and risks of Artificial Intelligence (AI) for countries in Africa? Broadly, AI is a class of technologies that enable machines to act with higher levels of intelligence to mimic human capabilities of sensing, comprehension, and action as well as the ability to learn from experience and adapt over time.

 AI is the most impactful technological revolution of our times and has the potential to disrupt and transform almost all aspects of human activities. Like other great technological breakthroughs, AI will boost productivity, thereby maximising output, and growth. The channels through which AI promotes productivity is that it represents an extra factor of production in addition to the traditional factors of capital, labour, and total factor productivity. In other words, the traditional growth model is transformed into an enhanced growth model that would now have four factors of production, viz: capital, labour, total factor productivity and AI.

Regarding the benefits to Africa, this note focuses on two sectors, viz: agriculture and financial services where these benefits are readily apparent. Despite the focus on these two sectors, it should be noted AI technologies will transform and benefit all sectors of the economy.

The agricultural sector is critical to Africa’s growth as it employs over 65 percent of the continent’s labour force, and accounts for 32 percent of Africa’s GDP. The World Bank estimates that African food markets will be worth USD 1 trillion (R15 trillion) by 2030 up from the current USD 300 billion (R4.5 trillion). AI technologies have the potential to improve productivity and efficiency at all stages of the agricultural value chain. These technologies can empower farmers to achieve higher crop yields as well as have greater price control and increase their income. For instance, drone technology can be more effectively and efficiently utilized to plant and fertilize seeds than utilizing human abilities. The 2017 Report of the Alliance for a Green Revolution in Africa observes that AI-powered analytics of crop data can identify diseases, enable the monitoring of soil health without the need of laboratory testing infrastructure, and facilitate the creation of virtual cooperatives to aggregate crop yields and broker better prices with suppliers. In South Africa there are growing AI agricultural start-ups such as Aerobotics, MySmartFarm, DroneClouds, among others that are spurring the technological revolution.

Another critical sector that is reaping the benefits of AI technologies is the financial sector. It is generally accepted that financial development is a key factor in promoting economic growth and reducing poverty in developing economies. In Africa, the objective is to ensure that 100 million more Africans can be financially included within the next decade. AI technologies have the potential to achieve this by changing the way Africans access financial services, save money, invest, and get insured. Recent studies suggest that approximately 40 percent of African banking customers prefer to use digital channels for transactions over branch channels. In the continent’s major banking markets that include Nigeria and South Africa, this share is significantly higher. For instance, in Nigeria, 59 percent of customers prefer digital, compared to 15 percent that favour branches. Machine learning is helping financial firms to track customer behaviour to offer tailored financial advice. For instance, smart machines can calculate and analyse credit scores and can be trained to track trading volatility and manage wealth and assets on behalf of investor customers.

South Africa spends heavily on AI-powered financial services. According to the South Africa Artificial Intelligence in Banking and Finance Industry Databook Series (2016-2025), AI spending has increased at 106.4 percent during 2018 to reach USD 69.0 million (R1.035 billion). Over the forecast period (2019-2025), spending on AI in South Africa is expected to record a CAGR of 35.4 percent, increasing from USD 119.2 million (R1.788 billion) in 2019 to reach USD 994.0 million (R14.910 billion) by 2025.

What are the challenges of AI? Alternatively, what are the risks inherent in AI?

Policymakers are at the centre in enabling development of AI technologies. They determine the legal, regulatory, and business environment. In this regard those who are forward-looking are working to enable AI and the digital economy through steps such as having clear legal and innovation-friendly data protection frameworks.

However, there are several barriers that will keep Africa lagging in the AI space. These range from lack of regulatory frameworks and supportive infrastructure, to lack of required skills and lack of quality data to fuel AI technologies.

Of concern is that AI technologies could be used by a few high-tech organizations to gather and analyse the data of individuals and make it available to many other actors who may engage in unintended or undesirable practices. Some of such practices could easily lead to discrimination based on age, gender, ethnic background, health conditions or social background. These may raise ethical concerns. For instance, using AI technologies, police forces can use advanced predictive analytics to predict a higher chance of crime rates in certain areas. This ability of AI to enable profiling of communities has potential to result in violations of human rights and legal principles.

Of equal concern is that AI applications in the migration sector have the potential of making migration unequal for some groups. In developed countries, algorithmic programs and machine learning have become integral to border management because of the commonly held assumption that they are less biased than direct human analysis. The European border and coast guard agency, Frontex, has set up risk-analysis cells across the African continent, the first of which appeared in Niger. This EU-Niger deal has been found to have risks of breaches due to Niger’s weak privacy laws. Furthermore, civil liberties and privacy groups have raised concerns that the use of AI technologies at US borders by the Department of Home Security could infringe on the human rights of foreign nationals.

Some scientists as well as businesspersons have expressed concerns that more advanced AI systems might in the future lead to humanity losing control. This has been termed the AI Control Problem whereby AI systems outperform humans within a narrow domain of application. Control theory (which stipulates that if one is not careful in specifying objectives in their full breadth, one risks generating unintended side effects) provides us with useful insights on what can go wrong. Greek mythology gives us an apt example. It is told that King Midas sought from the Greek god Dionysus powers that everything he touches should turn into gold. Since he failed to specify limitations to the god, he went home and touched his food and daughter who all turned into gold as per his wish that everything he touches should do so. The modern practical example is the revelation from the so-called “Facebook Files” in which Meta’s deep learning systems prioritized user engagement over other societal values to demonstrate the increasing control problem over advanced AI systems.

In conclusion, what we are observing is that the global rapid deployment of AI technologies taking place is moving faster than the legislative and frameworks that should regulate their usage. In Africa the AI regulatory environment is even patchier than it is in other regions of the world.


Africa Agriculture Status Report (2017). The Business of Smallholder Agriculture in Sub-Saharan Africa, Alliance for a Green Revolution in Africa (AGRA)

Korinek, Anton (2021). Why we need a new agency to regulate advanced artificial intelligence: Lessons on AI control from the Facebook Files. Brookings Report, December 8, 2021.

Thirtle, Colin, Lin, Lin and Piesse, Jenifer (2003). The Impact of Research-Led Agricultural Productivity Growth on Poverty Reduction in Africa, Asia, and Latin America World Development (Vol 31 Issue 12), 2003.

University of Pretoria (n.d). Artificial Intelligence for Africa: An Opportunity for Growth, Development, and Democratisation Access Partnership.