What strategy for a Made in Africa Artificial Intelligence?

Analysis

African governments, major tech players, and the academic world are betting on a "leapfrog" and an explosion of AI adoption across the continent by 2030. But is it really possible to establish a “Made in Africa” AI? Developer communities and research centers in Africa are growing rapidly, supported by national and international initiatives. However, the rise of AI will pose many challenges to the continent, such as its impact on employment, digital sovereignty, and adaptation to local contexts. An authentically African AI will need to overcome algorithmic biases and adapt to local contexts to succeed.

About

Marvin Amegashie is a consultant at Concerto. Specialising in online communication strategy, he advises companies on these issues. Contact Marvin at ma@concerto-pr.com for more information or to find out how Concerto can help you.

QUICK INSIGHTS

The potential impact of AI on employment, security, and sovereignty in Africa underscores the need for a cautious and inclusive approach.
National AI strategies in Morocco, Egypt, and Benin demonstrate the continent’s growing commitment to technological innovation.
For AI to be effective, it must be adapted to African realities, and the adoption and learning of this technology must be intensified.

Challenges and risks: A continent at a crossroads

Africa is particularly vulnerable to job displacement due to AI. According to McKinsey, 3.3 million jobs could be lost in South Africa alone due to AI by 2030. This risk is exemplified by companies like Teleperformance, whose operations in Egypt, Nigeria, and Madagascar are already threatened by AI. When Klarna announced the automation of certain tasks via AI, it led to a 21% drop in Teleperformance's stock, demonstrating the direct impact that AI adoption can have on jobs in Africa. Cybersecurity and data protection are also major challenges. Indeed, AI relies on large amounts of data, increasing the risk of breaches if not backed by robust cybersecurity measures. In Kenya, data protection failures have exposed millions of citizens, highlighting the need to strengthen cybersecurity regulations. Insufficient technological infrastructure poses another challenge: the continent's inability to access local cloud services. Many African startups depend on foreign cloud providers like Amazon Web Services (AWS), leading to high costs and consequently limiting their capacity to innovate. For example, the use of advanced AI models, such as Google's Gemini model, requires robust infrastructure and cloud capabilities that many local businesses cannot afford. This limits their development and innovation potential.

Emerging initiatives: Ambitious first steps

In response to these challenges, several African countries have launched ambitious strategies to develop AI. For instance, Benin is investing 4.68 billion CFA francs (7.15 million euros) over five years to integrate AI into healthcare, agriculture, and education. Morocco, in collaboration with UNESCO, is leveraging AI in education and public health, with the Mohammed VI Polytechnic University (UM6P) as a central hub for AI research in Africa. Through specialized programs and international partnerships, UM6P is training the next generation of researchers capable of tackling technological challenges. In Egypt, the national AI strategy launched in 2021 aims to achieve sustainable development goals and strengthen regional cooperation. The African Masters in Machine Intelligence (AMMI) program, for example, trains African students in AI to address the continent's specific challenges, such as improving food security through predictive systems for agriculture and optimizing healthcare services through AI-based diagnostic tools. Additionally, Google is collaborating with several African governments, particularly in Ghana, Kenya, and Nigeria, to develop AI solutions that improve weather forecasting and optimize the management of agricultural resources.

Towards an authentically African AI: Opportunities and perspectives

For Africa to fully benefit from AI, it must reflect local realities. The collection of data by foreign companies and algorithmic biases can exacerbate inequalities and threaten digital sovereignty, as these companies often use algorithms designed for different contexts, which can lead to biased or inadequate outcomes for African users. It is in this context that initiatives like Lelapa AI are crucial: They develop AI models adapted to African languages, allowing for better integration of local realities and offering more relevant solutions. Moreover, intensifying training is essential to enable young Africans to take ownership of these technologies. Despite the obstacles, local developers, often very young, are showing that technological barriers can be overcome through targeted initiatives. For example, the Moringa School in Kenya offers intensive training programs in web development and software engineering, with a growing focus on artificial intelligence. The integration of AI into local contexts also creates unique opportunities. Zipline, for instance, uses AI to optimize the delivery of medical supplies to remote areas using drones, demonstrating the transformative potential of this technology when adapted to the continent's needs. Africa is at a decisive moment: it must adopt AI to stimulate its development while managing the risks of increased inequalities, loss of digital sovereignty, and protection of personal data. For AI to realize its potential on the continent, it must be rooted in local specifics and supported by solid strategies. Thus, a 'Made in Africa' AI could not only be possible but also become a transformative force for the continent.