AI in Africa faces hurdles: These 5 countries are leading the way

While AI development in sub-Saharan Africa is hampered by poor infrastructure, lack of skills and regulatory confusion, a handful of countries in the region — spurred by the demand for economic development and recently, by the need to combat COVID-19 — are laying the groundwork for innovation in the technology,

Perceptions of Africa as lagging technologically have been highlighted once again, this time by the recent release of the 2020 Government AI Readiness Index published by UK-based Oxford Insights. Only Mauritius, South Africa, Seychelles, Rwanda and Senegal make it into the ranking, and the report paints a gloomy picture of Africa  “playing catch up in the area of artificial intelligence due to poor technological development, lack of infrastructure and a labour force that is not critical and innovative.”

While AI is widely seen as key to  innovation, there are varying levels of confidence in the ability to widely deploy the technology in certain areas of the continent. The five countries highlighted by Oxford Insights are taking different approaches in their efforts.

Getting it right, though, is not easy. Karthik Venkataraman, head of AI, Data and Intelligent Automation for Accenture in Africa, cautions that the businesses which have typically derived optimal value from AI are those that are digital by design, companies like Amazon, Facebook and Google. “For most other businesses though, the use of AI at an enterprise scale remains an aspirational topic,” he wrote in a recent Accenture report. Helping African businesses make this move requires savvy support from both industry and government.

Mauritius makes AI a national priority

Mauritius has built on its AI strategy launched in 2018, which spawned recommendations from the working group set out to research how AI could be integrated into national priorities and embedded in the business ecosystem. The island nation has created a Mauritius Artificial Intelligence Council, provides strong fiscal incentives for innovation-driven activities and has set up a National SME Incubator Scheme (NSIS). The country is the highest-ranked African nation on the Government AI Readiness Index at 45^th place globally.

Rwanda too are opting for a national strategy with the government looking to the Future Society, an independent, non-profit “think-and-do tank,” for its AI strategy development. “A comprehensive national AI policy, AI ethical guidelines, and a practical implementation strategy fit for the local context can serve as a powerful roadmap to achieve Rwanda’s national development and sustainable development goals,” the government said in a statement.

AI seems to make a regular appearance in success stories surrounding COVID-19 specifically, said Asif Valley, National Technology Officer at Microsoft South Africa. Countries that stand out include South Africa, where examples of AI readiness adoption include the SARS e-filing and mobile app and the use of AI and machine learning bots improving service delivery and engagement with citizens. The Department of Health, for example, adopted a WhatsApp-based chatbot to help and support people with COVID-related health queries and direct them to where they could access accurate health information.

Asif believes that the pandemic has opened the world’s eyes to the need for digital transformation as never before. “Remote learning, working and socialising has seen people embrace more technology than ever.” While many of the technologies stoking digital transformation have been around for a while, perhaps the most prominent is AI, Asif said.

AI to impact agriculture

He believes that one of the areas where AI can make the most impact in Africa, is is overcoming one of the most prominent challenges faced by the continent: providing food security for citizens.

AI in agriculture would see the use of advanced analytics and machine learning to bring centuries-old farming techniques into the modern age, giving farmers the tools to optimise crop yields and mitigate the effects of climate change. Given that agriculture sustains 70% of Africa’s livelihoods the impact of AI, IoT and edge computing to improve productivity and sustainability across the sector could be enormous.

But the challenges are not to be glossed over. Araba Sey, who wrote the Sub-Saharan African regional review to the Oxford Insights and IDRC 2020 AI Readiness Index, said that the region is lagging behind in AI because it is already lagging in other technological developments.

She adds that the “infrastructure is not in place; the labour force is not at critical mass to begin generating that kind of innovation internally and much of the population is in relative poverty.”

Better infrastructure is key to AI development

Infrastructure-wise, connectivity is a core challenge on the continent. Many countries are below the 20 percent critical mass necessary to achieve improved efficiencies and enhanced information flows for economic growth and innovation. “Consumer demand for wireless connectivity is surging and spectrum is a finite source. It is critical to intensively share underused spectrum bands,” Valley said.

The full benefits of cloud and AI cannot be realised for Africa unless everyone has access to reliable broadband connectivity. “We know there is an important correlation between broadband access and economic growth, and higher income, which is why there is a clear economic divide between those with broadband and those without. This divide will only expand as AI becomes more deeply embedded in everyday life,” Valley said.

Then there are issues like load shedding in South Africa and other parts of the continent, as electricity provision remains unstable. And political instabilities, both real and perceived, that can threaten potential partnerships and investments. While the regulatory landscape might not be the most obvious hurdle to adoption of AI, it is the most significant, Valley notes. The AI Readiness Index report notes that, “apart from the well-known infrastructure limitations, inconsistent policy and regulatory failure slow progress to AI readiness.”

A recent Microsoft EY study shows business leaders view regulatory requirements as the number one risk associated with AI. The same Microsoft EY research shows the majority of businesses in the Middle East and Africa are already piloting AI projects. Political will in places like Mauritius, Rwanda and South Africa has also seen movement to ensure regulatory support for AI development. South Africa’s President Cyril Ramaphosa has expressed government’s push for creating an environment in which the skills needed to develop and deploy emerging technology, including AI, are nurtured and developed.

Public-private partnerships address skills gap

Industry is also partnering with government to address skills needs. In July 2020, Microsoft partnered with the Central University of Technology (CUT), Free State to introduce its AI University Programme to address the demand for Artificial Intelligence skills in South Africa.

The programme — developed by Microsoft and delivered by its partner Gijima — teaches young multi-disciplined graduates with limited or no work experience to explore, transform, model, and visualise data, as well as to create the next generation of intelligent solutions. CUT, CUT Innovation Services (CUTis) and the Free State Premier’s office will be the first to benefit from the programme’s training, which has been designed to enable the skills transfer to the University lecturers who will lead the programme in the coming years. 

A focus on skills development is also apparent in Senegal where the first AI programming school in West Africa opened in Dakar in 2019. The inaugural class enrolled at the Dakar Institute of Technology (DIT) in September last year, attracting students from around the continent.

These developments hold promise but improving AI readiness will take time as well as intentional investment and public-private partnerships to make the use of the technologies work for all. Other countries in Africa can look to examples like Senegal, Mauritius, Rwanda and South Africa for models of how to kickstart AI strategy so that it becomes a priority and start laying the groundwork to leverage the benefits of AI that works for all.