How a Pan-African Roadmap Can Unlock the Potential of AI

Artificial intelligence (AI) promises to reshape all areas of modern life: the way we work and create, the delivery of public services, and how we guarantee our safety. But Africa risks being left behind in global efforts to both harness and regulate this transformative technology.

In our recent paper A New National Purpose: AI Promises a World-Leading Future of Britain we explored the United Kingdom’s potential in AI, enabled by the right policy framework. The European Union plans to invest €1 billion per year through the Horizon Europe and Digital Europe programmes. Likewise, many countries are moving to assure the safety of the technology: the EU is working on the first ever legal framework for AI’s safe use and the US has released its blueprint for an AI bill of rights.

But so far Africa has lagged behind, both in terms of the development of new technologies and the infrastructure required to harness AI. Indeed, in a recent report on AI readiness, the majority of sub-Saharan African countries ranked below the average score for the rest of the world.

AI in its present form risks exacerbating social and economic inequalities between those countries that play an active part in its development and those that do not. But Africa has an invaluable opportunity to harness AI and compete on the global stage. The continent has already proved its ability when it comes to technological developments, as seen with mobile money and internet access. It can do the same here, leveraging AI to serve its citizens and become a key player in the AI revolution.

The Tony Blair Institute for Global Change (TBI) has previously explored the steps that African countries should take for the strategic adoption of AI at a national level. However, one of the largest remaining barriers is the lack of a pan-African strategy. Countries are working in silos, with dispersed infrastructure, talent and capital. Here we lay out the state of play in Africa and explain why a pan-African solution is crucial – including key factors for the African Union to consider when creating a continent-wide enabling environment for AI.

The State of AI in Africa

The development of AI strategies across Africa varies considerably by country and region, and African countries are largely absent from most of the important global forums on the subject (see case study). It’s South Africa that sets the tone for the continent in terms of AI adoption, thanks to a strong ecosystem of technology hubs and research groups. Other countries – such as Benin, Egypt, Ghana, Nigeria, Senegal and Tunisia – have already developed or are developing national AI strategies, which often include recommendations for ethics, data privacy and security, and the development of human capital in AI. In addition, Côte d’Ivoire, Ghana, Morocco, Nigeria and Senegal have created strategies or roadmaps to promote the development and deployment of AI.

The variance of AI adoption between African countries highlights the need for cross-continent collaboration. Much of the continent lacks the infrastructure, compute power and human capital required for research and development; for example, while South Africa and Nigeria both have 80 or more innovation hubs, other countries have none. Plus, countries with resources are not sharing them, which limits the speed and scale of innovation; Morocco, which is home to the most powerful supercomputer on the continent, is a prime example.

A Pan-African Approach

A pan-African AI roadmap would result in AI systems that benefit all African countries, regardless of their level of development. This requires cooperation between governments, the private sector and civil-society organisations. Countries should also promote the advancement of AI by creating collaborative platforms, such as councils and working groups. These platforms could be used to share best practices, coordinate research and development, and promote collaboration among stakeholders; they can take inspiration from existing initiatives such as the European AI Alliance, AI4EU and the European Data Portal.

The AU has already begun work to foster this continental collaboration. It is developing a strategy for AI and, in 2019, launched the African Union Working Group on Artificial Intelligence, which is based on a proposal by Egypt to provide guidelines for the technology’s ethical development and use; it emphasises the importance of developing AI that aligns with African values and priorities, addressing unique challenges and opportunities. TBI recognises the merits of these initiatives but also seeks to build upon them: it’s possible to elevate existing frameworks and fill potential gaps by introducing elements such as ethical considerations, community engagement and inclusivity. This approach will ensure that the AI landscape in Africa evolves cohesively, drawing from past successes while propelling the continent toward a more dynamic and impactful future.

Several international initiatives also focus on supporting the responsible development of AI in Africa. One such is AI for Good, which promotes the development of AI technologies that are aligned with the UN Sustainable Development Goals. One of the major AI for Good projects in Africa is the FAIR Forward programme: implemented by German development agency Deutsche Gesellschaft für Internationale Zusammenarbeit in collaboration with various partners, its aim is to make AI more inclusive and sustainable via training and advice.

However, there is more to be done at a continental level. A comprehensive AI approach for Africa must be put together through a participatory process that involves input from a wide range of stakeholders. With that in mind, TBI proposes that the AU should build these six considerations into a pan-African strategy.

1. Data

Generative-AI projects require large amounts of high-quality data to train and fine-tune models. Africa has abundant demographic, health and economic data, but it is often fragmented and not easily accessible. As such, a comprehensive data strategy is necessary to enable the collection, sharing and analysis of data for AI development. Governments should work with the private sector and civil-society organisations to create data policies (such as the implementation of digital ID) that promote data sharing and collaboration while protecting individual privacy.

In an important step, the AU published its Data Policy Framework in 2022. This sets out a shared vision and recommendations to steer AU member states towards creating adequate national and regional data frameworks, with the aim of unlocking the power of an integrated African data ecosystem through a common approach. African governments must now implement these recommendations by establishing national data policies and leveraging the African Continental Free Trade Area (AfCFTA) phase III negotiations; this will harmonise data-protection frameworks and allow for the development of a data-inclusive AfCFTA implementation plan.

2. Infrastructure

To support AI development and deployment, and enable the data sharing outlined above, Africa needs to prioritise – and attract – investment in data infrastructure: broadband networks, supercomputing clusters and graphics-processing units, data centres, servers and cloud computing optimised for machine learning. And according to a new report from the African Data Centres Association and Xalam Analytics, another 1,000 megawatts of power and 700 new facilities are needed to bring the rest of the continent up to speed with South Africa.

Currently there are very few supercomputing projects in Africa. South Africa and Morocco have developed clusters of them but, due to the high cost of large-scale compute systems, a pan-African project is needed to realise the full potential. The AU could take inspiration from the recently launched pan-EU supercomputing initiative.

3. Education and Workforce Development

To ensure that Africa has the necessary expertise to deploy AI systems, investment in education and skills development should be prioritised. This could include initiatives to develop a talent pipeline of data scientists, machine-learning engineers and other AI professionals, as well as AI knowledge-sharing platforms.

Rwanda’s African Centre of Excellence in Data Science focuses on linking academic and industrial research on deep learning; the discovery of young talent is also encouraged. There are similar centres in Ethiopia, Ghana, Malawi, Morocco and South Africa, all of which seek to accelerate innovative research in new technology, including AI, and support local technology ecosystems. Meanwhile, Africa’s first AI research centre was launched in Congo-Brazzaville, in 2022.

4. Unlocking Capital

The investment required to develop a generative-AI project can range from tens of thousands to several million dollars. Public-private partnerships should be encouraged; investment can be made attractive through tax incentives and other policy mechanisms, such as friendly terms for universities spinning out research and reduced bureaucracy in grant applications. Governments should also prioritise investment in AI, including funding for universities and research institutes, as well as at the intersection of AI and other emerging industries (such as life sciences).

5. Collaboration

International collaboration should be facilitated between African countries and global AI leaders such as China, the US and the EU. The GPAI provides an excellent forum for global collaboration, and there are interesting prospects for more African countries to join this platform, following Senegal’s example. In addition, the Global Summit on AI Safety, due to be held in the UK later this year, has the potential to create international organisations that will develop consensus and insights. African countries should work to develop tech-forward foreign policy so that they can ensure their place at the table.

6. Governance and Regulation

Research, development and deployment must be guided by a robust regulatory framework that balances innovation with accountability. The Malabo Convention, a framework for preventing cybercrime on the continent, is now in force. However, many countries in Africa have a significant cyber-security maturity gap, and building on weak infrastructure has the potential to cause larger problems.

To grow AI capabilities, African policymakers must prioritise regulations that address issues such as bias, transparency and accountability. These steps should be taken in consultation with the private sector, civil-society organisations and other stakeholders. The AU should also consider encouraging the secondment of leading AI experts to governments, as seen in countries such as the UK, to enhance understanding, expertise and policymaking capabilities in the AI domain. These experts could come from within Africa, as well as the continent’s diaspora.

We are at a pivotal moment in the development of AI, and the opportunity for African countries cannot be underestimated – but the ambition shouldn’t be to merely catch up.

Through its collaborative approach, a pan-African strategy acknowledges the value of ongoing efforts, building upon the progress already made in countries such as South Africa and Morocco, and by existing initiatives. Africa can harness its collective strength to ensure a seamless transition to a unified AI development path – and in turn, the continent can establish itself as a major global player.

Lead image: Getty