AI in West Africa: Sober audit of current landscape

The conversation on Artificial Intelligence in Africa largely falls into two camps. One faction believes the continent cannot afford to be left behind and must focus all resources on becoming up to speed with AI advancement. The other thinks it a distraction, willing the continent’s efforts be focused on more existential problems like hunger, electricity, and health.

Join the Premium Times WhatsApp Community For Quick Access To News and Happenings Around You.

Open in WhatsApp

While I admit that our more pressing problems are indeed pressing, I am inclined to tilt more on the side of those championing that the continent focuses on AI. However, it cannot be a zero-sum game. We can build AI capabilities while also solving core problems; in fact, there is no other way to do AI.

As Ojoma Ochai argues, the same was said for mobile communication in the 90s. The problems remain largely the same, and yet mobile phone penetration on the continent stands at a significant level. The larger question here is: can and should Africa attempt to leapfrog?

Nigeria leads, but the continental picture remains complex

Nigeria has emerged as West Africa’s undisputed AI leader, boasting over 400 AI firms and reclaiming its position as Africa’s top venture capital destination with $520 million in equity funding in 2024. The country’s National AI Strategy, launched in August 2024, positions Nigeria as “West Africa’s AI hub and model for Africa’s economic growth,” developed through collaboration with 120 global AI experts.

Stay Ahead with Premium Times

Add as a preferred source on Google

This progress is tangible. A Nigerian startup, Ubenwa Health, is using AI to analyse baby cries for birth asphyxia detection with 96 per cent accuracy, while Kudi leverages conversational AI for electronic banking services, raising $5.9 million from Y Combinator. The government has backed this ecosystem with a $1.5 million AI fund and partnerships with Google for an additional ₦100 million startup fund.

Yet, context matters. Sub-Saharan Africa still has the lowest average score on the 2023 government AI readiness index by Oxford Insights, a mere 30, compared to North America’s 81. Within Africa, Morocco (0.56), Egypt (0.53), and Tunisia (0.47) significantly outperform most Sub-Saharan nations in AI preparedness. Even the situation in Ghana, which hosts Google’s African AI research centre, and Senegal, committed to training 90,000 citizens in data science by 2028, highlights how much ground West Africa must cover.

Continental initiatives signal serious intent

Several African countries have taken significant steps towards AI adoption. Rwanda leads the charge, having approved a National AI Policy in April 2023, while Mauritius has maintained a national AI strategy since 2018. Egypt unveiled its comprehensive National AI Strategy in July 2021, and South Africa launched its Draft National AI Plan in April 2024, aiming to attract ZAR70 billion (approximately $3.7 billion) in investment by 2030.

Most significantly, the African Union endorsed its Continental AI Strategy in July 2024, marking unprecedented pan-African coordination. This Africa-centric, development-focused approach emphasises continental collaboration over fragmented national efforts, a crucial recognition that individual African states may lack sufficient resources for separate, siloed strategies.

The ecosystem shows promise. Africa houses over 2,400 companies specialising in AI, with 41 per cent being startups that have attracted approximately $2.02 billion in investment. However, 63 per cent of African AI startups remain in initial and intermediate development phases, while around 50 per cent are concentrated in just seven countries, underscoring both the opportunity and the uneven distribution of AI development.

Infrastructure realities demand urgent attention

Africa’s infrastructure constraints create fundamental barriers to AI deployment. The continent holds less than 1 per cent of global data centre capacity, with approximately 100 data centres concentrated mainly in South Africa, Nigeria, and Kenya. Only a third of Africa’s 80-plus cities with populations exceeding one million can boast at least one Tier III data centre facility.

The energy challenge is particularly stark. While traditional data centres typically use 5 to 10 kilowatts per rack per hour, AI centres can consume up to 50kw per rack per hour. To put this in perspective, a single rack in an AI data centre would require over eight times more electricity in a day than an average Nigerian consumes in an entire year. This comparison underscores the magnitude of infrastructure challenges when nearly half of sub-Saharan Africa lacks access to electricity.

Digital connectivity compounds these constraints. The region’s 27 per cent mobile internet connectivity rate significantly trails the global average of 57 per cent, with data costs reaching 8.76 per cent of average income in Africa versus 1.54 per cent in Asia. As recent Science journal analysis notes, African businesses struggle with electricity supply that is “among the most expensive and unreliable on earth.”

The data and talent paradox

Africa grapples with what Shanta Devarajan of the World Bank terms a ‘statistical tragedy’, data scarcity stemming from underfunded national statistical systems, political interference, and resource constraints. This poses a fundamental challenge for AI development, which relies heavily on vast amounts of quality data.

The talent picture presents both challenges and opportunities. Many tertiary institutions across Africa still rely on outdated, overly theoretical digital technology curricula, failing to produce graduates ready for the demands of the digital workforce. The current AI workforce includes a growing pool of data annotators, often the entry point for AI work, but these workers frequently face poor labour conditions, including underpayment and exposure to potentially harmful content.

ALSO READ: The role of artificial intelligence in healthcare: Innovations and impact

However, demographics favour transformation. An estimated 95 per cent of youth aged 15-24 in Africa are engaged in the informal sector, representing both challenge and opportunity. With proper training and support, this large pool of young talent could fuel Africa’s AI sector growth. Initiatives like Nigeria’s 3 Million Technical Talent programme and the African Development Bank-Intel partnership to train 3 million Africans in AI skills represent the scale of intervention required.

Success stories prove leapfrogging is possible

Despite constraints, breakthrough achievements demonstrate Africa’s AI potential. InstaDeep’s $684 million acquisition by BioNTech in 2023 represents Africa’s largest AI exit to date, proving African AI companies can compete globally. The Tunis-based company, with operations across West Africa, collaborates with Google AI on locust detection systems for agricultural protection, directly addressing local challenges.

Google’s AI research centre in Ghana has produced tangible regional benefits: flood forecasting systems for West and Central Africa, satellite imagery analysis that quadrupled African buildings on Google Maps, and machine learning for Google Translate supporting 24 new languages, 10 of which are African. These initiatives demonstrate how international partnerships can drive local innovation while addressing regional challenges.

Similarly, West Africa’s successful mobile money adoption proved the region’s capacity to bypass traditional infrastructure development paths. The demographic dividend is substantial: Africa will house half the world’s working-age population by 2063, with high potential for technology adoption among its youthful population.

The window is narrowing

Africa finds itself at a crucial juncture in the global AI race. Projections suggest AI could contribute $1.2 billion to the continent’s GDP by 2030, provided it captures just 10 per cent of the rapidly growing global AI market. This transformative potential comes with urgency, as Bosun Tijani, Nigeria’s Minister of Communications, warns: “We would be failing our people and future generations if we ignore artificial intelligence because these technologies will shape what you think, how you think, and how you operate.”

The challenges are formidable: limited infrastructure and investment, data scarcity and quality issues, shortage of skilled talent, low market awareness, and significant regulatory gaps. Yet opportunities abound: a large, largely untapped market for AI solutions, potential for leapfrogging traditional technologies, context-specific application development, and a growing pool of young, tech-savvy talent.

While Europe focuses on governance frameworks like the EU AI Act, and America and Asia lead in innovation and infrastructure, Africa’s role in the global AI landscape is still evolving. By focusing on building its ecosystem and developing contextually relevant solutions, Africa could eventually play a larger role in both innovation and governance of AI technologies.

The key questions remain: What does Africa-centric Artificial Intelligence look like? And how do we sustainably create it? As the brilliant Abraham Augustine succinctly puts it, “Leapfrog away. As long as you look before, during, and after you leap.” This encapsulates the balanced approach Africa must take, ambitious yet cautious, innovative yet grounded in local realities.

The continent can neither afford to ignore its existential problems nor watch the AI revolution unfold from the sidelines. There is a small and closing window of opportunity for African nations to become serious players in this field, and the future of the continent may well depend on staking a claim in AI development. The journey ahead is long and fraught with challenges, but the first steps have been taken.