Healthy Appetite for AI in Africa, but Risks Remain


The first AI Summit in Africa was held in Cape Town in South Africa during the second week of November. Tractica chaired the proceedings of the 2-day AI conference and presented on the African AI opportunity. The AI Summit Cape Town was part of the larger AfricaCom event, which is an annual gathering of the African telecom ecosystem, but looks at the broader information and communications technology (ICT) landscape in Africa, covering areas like the Internet of Things (IoT), mission critical technologies, smart cities, digital media, fintech, and digital transformation of the enterprise sector.

In the run up to the event, Tractica laid out the challenges and opportunities for AI in Africa in the white paper entitled, “Artificial Intelligence Opportunities in Africa: Creating Conditions for Success”, the highlights of which I presented on the second day of the AI Summit. However, it was good to be on the ground in Africa, talking with AI enthusiasts and stakeholders to better understand how Africa views AI and how AI can help change the lives of ordinary people.

Going into the conference, I understood that Africa should not be viewed as a monolithic region, but a mix of multiple regions, countries, and cultures, and that like AI, which is anything but monolithic, there are no golden rules or singular approaches. For the most part, I came away reinforcing that belief, but was also surprised that there is more AI activity in Africa than what we hear in the media. In addition, I came away feeling that Africa needs better representation on the global stage, especially when looking at the social, ethical, and safety issues related to AI.

AI was a key theme on the main AfricaCom Headliners stage with mentions by telecom operators like MTN and Orange and in relation to panel discussions around public policy, smart cities, and digital banking. The AI Summit drilled down deeper into some of these topics, with several use case-focused presentations covering the opportunity ahead, rather than existing implementations. These included:

  • AI in customer-facing banking services by Stanbic Bank
  • AI in service provider promotions and value-added services by Velti
  • AI in legal services by Webber Wentzel
  • AI in retina scanning by Retina-AI
  • AI for mobile-driven digital identity by Element
  • AI used in drone-based diseases detection for agriculture by Aerobotics

There was also a great presentation by Jacques Ludik, who heads the Machine Intelligence Institute of Africa (MIIA), covering all the work that it is doing with other partner organizations in promoting AI and exploring use cases on the African continent across healthcare, finance, agriculture, education, energy, and government, among others. On the show floor at AfricaCom in the AHUB startup section, I also noticed a number of AI startups taking part in incubators like Venture Capital for Africa (VC4A), with agriculture being a key area where AI has potential in Africa. When speaking with delegates at the conference, I also learned about AI activity by startups and data science enthusiasts tackling one of the worst droughts in Cape Town making use of the Cape Town Open Data Portal. Amazon Web Services (AWS) sees the hunger for AI in Africa and is known to be opening its first Africa data centers in Cape Town by 2020 and is on a hiring spree with AI and machine learning as a key focus area.

But there was a broader theme that emerged around the need for socially inclusive and ethical AI borne out of a keynote presentation by Nanjira Sambuli of the Web Foundation, followed by a panel discussion on the topic. The discussion touched on key issues like algorithmic bias, gender and race bias in the AI workforce, and the impact of AI on jobs. At one point during the discussion, there was a heated debate about AI regulation, with some in the audience preferring to leave AI unregulated in Africa, letting the market and government deal with the jobs and automation aftermath.

The stark differences of opinion in the regulation of AI is a reminder that there are major risks for AI in Africa, where unchecked applications of AI could cause major social and political problems. In a European, U.S., or Chinese context, we are already hearing the need for AI safety and regulation and if there is one region of the world where the risks of AI need to be discussed and better understood, it is in Africa. Africa has the youngest population in the world, with 200 million people between the ages of 15 and 24, but more worryingly, 60% of the continent’s youth is unemployed, which has a direct correlation with crime and violence.

If AI is applied unchecked and unregulated, it is very likely that the youth dividend in Africa will become a burden, with AI destroying jobs faster than what can be created. The focus on skills development in Africa came up time and again during the conference, and most agreed that AI skills should be the number one priority for private companies and public policy makers. There was a specific suggestion at the AI Summit to have a separate and dedicated track around the techno-social impact of AI in Africa, which is a great idea for next year’s AI Summit in Cape Town.

Overall, there is tremendous potential for AI in Africa. There is a healthy appetite and lots of activity on the ground, with several AI and data science professionals working on using AI to solve problems in the areas of healthcare, agriculture, public sector, finance, transportation, and many others. The lack of digital maturity in the enterprise sector is a problem in Africa, which is why we do not see a lot of real-world implementations of AI today. There is a broader risk of AI being implemented without proper AI safety, ethics, and bias checks, but unless African business and government leaders get on board, it is unlikely that we will see progress. The good news is that conversations have begun and conferences like the AI Summit in Cape Town are helping to move the needle.

Comments are closed.