Mobile-enabled edge computing recognized as key solution to overcome significant infrastructure challenges
LondonAfrica currently accounts for just 2.5% of the global artificial intelligence (AI) market, but emerging applications could boost the continent's economic growth by $2.9 trillion by 2030, according to AI4D Africa. A new GSMA report,AI for Africa: Use cases that make a differenceDeveloped from existing research and interviews with leaders from civil society, NGOs, academia and the private sector, it identifies over 90 AI use case applications in cutting edge technology markets (Kenya, Nigeria and South Africa) that have the potential to have socio-economic and climate impacts.
However, to unlock AI's potential, significant barriers must be overcome, such as limited availability of data centers and expensive technology investments. By addressing the digital skills gap and increasing the availability of smartphones, mobile-based AI solutions could be a practical way to circumvent current limitations and unlock the full potential of AI across the continent.
Of the currently identified AI use case applications in Africa, the majority relate to agriculture (49%), climate action (26%) and energy (24%).
Agriculture use cases drive mainstream AI adoption
Agriculture employs 52% of Africa's workforce and accounts for 17% of GDP on average. In sub-Saharan Africa, up to 80% of food is produced by smallholder farmers, who often use traditional techniques and lack access to information that can help improve yields. The GSMA found that the majority of AI use cases in agriculture are machine learning (ML)-enabled digital advisory services that provide farmers with data-driven advice to adopt climate-smart agricultural practices and optimize productivity. These solutions typically reach farmers via mobile devices, highlighting the importance of device ownership, digital skills and literacy, and ease of use.
Providing affordable and reliable energy services
The region faces significant challenges when it comes to energy access and reliability, with half of the population living without electricity. Today, AI-enabled solutions in Africa are improving both on-grid infrastructure and off-grid systems, with use cases such as predictive maintenance, smart energy management, energy access assessment and productive use financing, monitoring and expanding services in energy-deficit areas. The GSMA highlights that improving energy access and efficiency in the region is essential, as it creates a virtuous cycle through enhanced use of the internet and digital tools, cellular networks and broadband, and the generation, transmission and distribution of data needed for AI capabilities.
Supporting climate change measures
Despite accounting for less than 3% of global energy-related CO2 emissions, Africa is disproportionately affected by climate change. Without intervention, climate-related emergencies could reduce Africa's GDP by 8% by 2050. The GSMA says the increased availability of remote sensing technologies and satellite imagery has spurred the development of natural resource management use cases, where AI is being used to monitor biodiversity and protect wildlife. Early warning systems that provide predictive analytics and real-time disaster assessments to provide timely warning of climate emergencies and other natural hazards are also already being improved by ML models, significantly improving forecasts in data-scarce regions.
The vast majority (98%) of AI use cases in Africa are categorised as predictive AI applications leveraging ML approaches due to the availability of historical datasets, ease of application and lower computational requirements compared to generative AI models. The GSMA has identified several hurdles that must be overcome to realise the full potential of AI opportunities, including more nascent use cases and generative AI, which are key to driving long-term socio-economic benefits.
Unleashing Africa's AI capabilities
Extensive, diverse, and representative data is essential to effectively train AI models. It is important that datasets reflect the complexities and nuances of African markets, rather than mimicking data from the global North. For example, there are currently significant gaps in the availability of local language data across Africa. Despite efforts by governments and the private sector, high-quality, locally relevant data is very limited and difficult to access, hindering the development and scaling of AI.
AI development also requires robust infrastructure and computing power. As AI applications expand, data center energy demands and hardware and software costs will rise. Africa already has a shortage of data centers, and countries such as South Africa and Kenya face prohibitive costs for graphics processing units (GPUs), accounting for 22% and 75% of GDP per capita, respectively, making them significantly more expensive than high-income countries.
As local computing ecosystems grow, countries can leverage mobile-first markets to develop distributed or hyper-local edge computing, where tasks are performed on devices such as phones and laptops, reducing reliance on high-performance data centers. After the underlying models are trained on large datasets, the AI models can be transferred to smartphones for fine-tuning. With smartphone penetration at 51% and expected to reach 88% by 2030, mobile-based edge computing will be central to expanding AI adoption and capabilities in Africa.
Max Cuvelier-Giacomelli, Head of Mobile Development, GSMA said: To harness the transformative potential of AI across Africa, we need to focus on improving the skills of both AI developers and users, especially among underserved populations. Better training programs are essential, especially now that we face a global brain drain of AI talent. To ensure Africa is not left behind, we need strong partnerships across a broad ecosystem of partners, including “Big Tech”, NGOs, governments, and mobile operators. Policies must also evolve to address inequalities, ethics, and human rights issues in AI deployment. As African countries develop their own AI strategies, active participation in global forums will be crucial to define regulatory frameworks that foster ethical AI development and protect societal benefits, moving towards sustainable solutions that benefit all African communities.
