Envisioning Africa's AI governance landscape in 2024

#
Christina Morillo via Pexels

Authors

African voices and perspectives are important in policy dialogues around artificial intelligence (AI). Melody Musoni explores the anticipated developments shaping AI governance in Africa, with the adoption of an African strategy on AI as a major policy milestone.

% Complete

    Summary


    African voices and perspectives are important in policy dialogues around artificial intelligence (AI). With the Global North leading in AI innovation and policy discussions, developments in Africa are often overlooked or seen as of lesser importance. 

    In this policy brief, I explore the anticipated developments shaping AI governance in Africa, with the adoption of an African strategy on AI as a major policy milestone. A continental strategy on AI has the potential to clearly articulate how African countries will approach AI governance and the important principles which will inform and shape AI governance in Africa. 

    Importantly, African participation in global AI governance should not be out of tokenism but, rather, African voices should have an impact in ensuring that AI promotes socio-economic development in Africa. Overall, Africa still needs the support of the private sector and international partners to build its own AI innovation industry.

    African participation in global AI governance should not be out of tokenism but, rather, African voices should have an impact in ensuring that AI promotes socio-economic development in Africa.

    Introduction 


    If 2023 was the year in which discussions around AI governance took global centre stage, then 2024 is poised to witness the translation of these deliberations into concrete actionable policy measures. But what do these global policy discussions actually mean for the African data economy? How should Africa strategically shape its AI policy in a manner that creates value and promotes social and economic development for Africans? These questions underscore the imperative for African countries to proactively govern data, the data value chain, AI systems and the entire AI lifecycle. This makes 2024 a very critical year for African governments to implement continental commitments around data, data protection, continental free trade and cybersecurity. In this brief, I explore anticipated developments in 2024 and present recommendations policymakers need to implement if Africa’s vision of being a global actor in AI is to become a reality. 

    What should we expect in 2024?


    1. Endorsement and adoption of an African continental AI strategy


    In 2023, the African Union (AU) developed a Continental Strategy on AI (the AU AI Strategy). While the strategy is not yet publicly available (AUDA-NEPAD 2023), plans are underway for its endorsement at the upcoming AU Summit, scheduled for February 2024. We can expect that this AU AI Strategy will demonstrate leadership by the AU and provide a continental vision that outlines ‘the African approach to AI governance’. The strategy will likely play a pivotal role in articulating Africa’s distinctive vision for Al, but a lot of this depends on how it is framed. In this sense, there are a few areas for the strategy to address. Some of the major concerns levelled against AI in Africa are the limited use of African datasets for proper training and application of AI applications and the overreliance on algorithms produced outside the continent. This results in Africans using biased AI which does not reflect or represent their African contexts and is likely to reinforce different forms of historical and societal biases such as racial and gender bias (Erastus 2021). Similarly, global AI discussions have so far marginalised and excluded African contexts, cultures, and values such as the African value of ubuntu (Gwagwa et al. 2022), a value which promotes the importance of community over individuality.

    We can expect that the strategy will clearly articulate how the AU will address the underlying challenges of limited use of African data and provide a clear policy strategy on how the AU intends to address algorithmic biases moving forward. This depends on the strategy clearly articulating the continent's priorities and underlying principles that should guide AI development, adoption, and use. If Africa wants to communicate to the global community on what an African approach to AI entails, then the strategy should emphasise the importance of respecting, protecting, and upholding African cultural practices and values by laying out key AI principles applicable and relevant for Africa. Overall, I anticipate that the AU AI Strategy will elaborate on respecting, protecting and upholding human rights and freedoms (including the use of AI in mass surveillance), complying with international law, and providing solutions which lead to inclusive social prosperity for African people and African communities and boosting the African data economy. To be more impactful, the strategy should have a mechanism for the AU to support member states to implement this strategy at the national level to promote policy coherence and coordination.

    We can expect that the strategy will clearly articulate how the AU will address the underlying challenges of limited use of African data and provide a clear policy strategy on how the AU intends to address algorithmic biases moving forward.

    2. Increase in the number of national AI strategies and/or policies


    Currently, less than 10 African countries have national AI strategies and/or national AI policies. Notable examples of proactive countries in this regard include Rwanda, Benin, Egypt, Morocco, Mauritius, Tunisia, Sierra Leone, and Senegal. These countries have demonstrated progress in formulating their respective AI and data policy frameworks, with shared objectives focusing on the necessity for increased research and development in AI, construction of robust data infrastructures, skills development, developing local AI solutions, increasing AI funding and investment incentives, providing secure and trusted African data sets, et cetera. The collaborative efforts of various international players such as the World Economic Forum, the Future Society, UNESCO, and Team Europe (GIZ FAIR Forward and AU-EU D4D Hub), have played a crucial role in supporting these African countries in their AI governance.

    Once the AU AI Strategy is adopted, African countries should use the strategy as a blueprint to develop their national AI strategies in a way that promotes policy coherence and coordination. Some countries are moving at a relatively faster pace towards AI governance and developing their national AI strategies and policies. The Southern African Development Community (SADC) Regional Economic Community (REC) member states, especially Namibia and South Africa, are likely to take more concrete steps to fulfil their commitment to developing AI policies and regulations to safeguard against the use of high-risk AI as outlined in the Windhoek Statement on AI (UNESCO and Government of Namibia 2022). South Africa is likely to lead the pack in taking decisive steps to develop a clear national AI strategy (Government of South Africa 2020). The country has already been implementing some of the recommendations it received from a team of AI and 4IR experts including establishing an AI Institute and investing in human capital (Government of South Africa 2020). Developing a clear AI policy seems to be the next step of the process but policymakers may need to prioritise and finalise the draft Cloud and Data Policy. South Africa also worked with the Smart Africa Alliance to develop the Smart Africa Blueprint on AI, which can inform and shape the development of South Africa’s national AI strategy. With the upcoming elections in the country, there may be a fast-tracking of these efforts as part of the ruling party’s campaign to demonstrate their efforts in digital work. When compared to other sub-Saharan countries, South Africa has made notable advancements in data governance and building of digital infrastructure, securing the second position in the Government AI Index for sub-Saharan Africa, just behind Mauritius (Oxford Insights 2023). Meanwhile, Namibia has initiated discussions with UNESCO on the implementation of the UNESCO Recommendation on AI (Terblanché 2023).

    Within the Economic Community of West African States (ECOWAS) REC, countries to follow include Côte D’Ivoire, which is also working on the implementation of the UNESCO Recommendation on AI, and Nigeria (DigiWatch 2023), which has started working on developing its AI strategy. In the East African Community (EAC) REC, Kenya (Indeje 2023) and Ethiopia (Fana News 2023) are in serious discussions around developing AI policies. There is an opportunity for Africa’s international partners like the European Union to continue to support this work.

    3. Strengthening of robust data protection laws


    In 2024, the AU and RECs need to prioritise supporting the remaining 15 countries with no data protection laws to adopt such laws (Musoni et al. 2023). Compliance with data protection is still a major challenge in Africa. Most African countries still require technical and financial support in order to operationalise, implement and enforce their data protection laws. Significantly, African governments need to strengthen the institutional capacities of data regulators, enabling them to play a more proactive role in holding AI actors accountable. AI actors need to ensure that they are accountable for the design and implementation of AI systems in such a way as to ensure that personal data is protected throughout the life cycle of the AI system. In instances where there is unlawful processing of personal data within the AI system, data regulators should exercise the appropriate oversight to ensure accountability of AI systems. Given that a significant portion of AI utilised in Africa is still imported, governments must urgently consider shaping legal frameworks that hold all AI actors accountable throughout the entire AI lifecycle. In developing these frameworks, African governments should develop effective mechanisms to hold foreign AI actors accountable. In instances where there is a need to track down AI actors, this may require getting support from the countries that those companies come from. Of course, this is going to be a challenge and is another reason to support and use locally developed AI solutions. 

    Due to the complexity of AI systems, African data regulators should consider the recruitment or engagement of experts in AI and data science with the requisite technical expertise to support regulatory efforts. Some of the functions of the experts would be to support data regulators in conducting privacy impact assessments and AI ethical impact assessments to understand the risks that these technologies have on people’s fundamental rights and freedoms. Data regulators across Africa should start thinking about ways to attract and retain such experts in the face of competition from the private sector which often offers competitive salary packages. It might make sense for data regulators to pool resources and find AI experts who can support work across several African jurisdictions. Institutional capacity should also be extended to courts and judicial services.

    4. Influence from Brussels and the push to legislate


    The coming into force of the European Union AI Act (EC 2021) will shape AI governance in Africa, directly or indirectly. This law applies to providers placing AI systems into the EU market irrespective of where the provider is established or if the providers and users of AI systems are located in a third country, where the output produced by the AI system is used in the EU. The implications of the EU AI Act’s extraterritorial application is that any African AI companies targeting European markets will be bound by this law. We previously witnessed how the EU GDPR influenced African data protection laws through what Bradford titled ‘the Brussels Effect’ (Bradford 2019). It is probable that there might be a repetition of the Brussels effect within the AI governance space where AI companies in Africa and multinational corporations will lobby African governments to develop laws which mirror the EU AI Act to make it easier for them to navigate both African and European markets. The EU AI Act adopts a human rights-centred approach where fundamental human rights are respected when using AI and provides for human oversight throughout the AI life cycle. The EU AI Act can be a useful reference point for African governments in crafting their own national AI laws which align with the AU vision, national laws and international law. But it remains to be seen whether, and to what extent, African governments will borrow text from the EU AI Act. 

    Beyond the influence of Brussels, there are local factors influencing the heightened focus on AI legislation by African governments. African governments are experiencing the adverse consequences of unregulated AI, particularly in the realms of cybersecurity, national security interests and democracy (Vainilavičius
    2023). The urgency to regulate AI is expected to intensify, especially given the imminent elections in nearly 20 African countries in 2024. AI can potentially be used in disseminating misinformation, fake news, and disinformation which threatens the credibility of democratic processes. These risks and concerns are likely to prompt governments to prioritise the development of guidelines specifically addressing AI in electoral contexts. In our previous work, we proposed that African governments should not rush to regulate AI, but should establish AI regulatory sandboxes as a preliminary step (Musoni 2023). We should expect to see the mushrooming of several AI regulatory sandboxes across the African continent. 

    Beyond the influence of Brussels, there are local factors influencing the heightened focus on AI legislation by African governments. African governments are experiencing the adverse consequences of unregulated AI, particularly in the realms of cybersecurity, national security interests and democracy.

    5. Increased African participation in global AI governance


    African voices have previously been excluded in global discussions on AI at various international fora. The effect of this exclusion was that African interests and African contexts were not seriously considered when developing AI. In 2023, we saw an increase in the number of African stakeholders participating in discussions at international fora, as well as an increase in the number of African experts forming part of the United Nations AI Advisory body (UN 2023). The G20 has also been contributing towards the global AI policy discussions, adopting the G20 AI Principles which were drawn from the OECD Recommendation on AI and has repeatedly emphasised the development of ethical and responsible AI. During the Indian G20 Presidency, the AU was welcomed as a G20 member and during the 2023 India Summit (Government of India 2023), the G20 leaders reaffirmed their commitment to G20 AI Principles and committed to govern AI in a way which promotes innovation while achieving Sustainable Development Goals. With the AU’s new role within the G20, there is an opportunity for Africa to contribute towards the global AI policy discussions. Africa can also keep pushing and advocating for its position through South Africa which will be hosting the 2025 G20 Summit. With the consecutive G20 presidencies being led by the Global South (India 2023, Brazil 2024, and South Africa 2025), there is an opportunity for African voices and Global South voices on AI to be heard. 

    African countries and African experts must continue participating in multi-stakeholder dialogues such as the Internet Governance Forum or the Global Partnership on AI to learn from countries with more experience in AI governance and share African experiences and expectations. African countries should make sure that the participation of African experts in international fora is not out of tokenism but rather, African views and Africa’s position on AI are considered in global frameworks on AI which will lead to actionable steps and concrete solutions that promote social and economic development of African people. 

    6. Booming of an African innovation system


    African governments understand the value of building digital infrastructure and developing local AI solutions to address African challenges and assert their AI sovereignty interests. As such, some governments are encouraging investment in AI research and innovation by providing research funding, and domestic resource mobilisation especially through engaging the private sector to invest, forming Public Private Partnerships on digital projects, and negotiating with international partners. AI-specific investment projects include Rwanda’s Seed Investment Fund which was created under Rwanda’s AI Strategy to promote an enabling environment for the government to co-invest alongside angel and venture capital investors in AI companies (Government of Rwanda 2023). Egypt’s AI Strategy is geared to increase funding offered to AI start-ups, creating AI-start-up incubators, and providing incentives for companies to purchase AI products locally instead of importing (Egypt National Council for Artificial Intelligence 2021). The Mauritius AI Strategy has different innovation and commercialisation schemes which provide fiscal incentives to different AI actors including research institutions, academia, private companies, start-ups and innovation hubs (Mauritius Working Group on AI 2018). 

    While some African governments are providing fiscal incentives and financing, these investment packages are not enough. The majority of investments and funding is likely to come from international partners, international organisations and the private sector. African countries should take advantage of international partnerships and initiatives, such as the EU’s Global Gateway, to direct investments towards the kinds of infrastructure and investments necessary to grow Africa’s innovation ecosystem. What is critical for African countries is to create a conducive enabling environment for research and innovation opportunities to attract more AI actors and investment opportunities. A successful AI investment agenda could lead to a boom of African AI start-ups and innovation hubs in the coming years.

    African countries should take advantage of international partnerships and initiatives, such as the EU’s Global Gateway, to direct investments towards the kinds of infrastructure and investments necessary to grow Africa’s innovation ecosystem.

    7. Increase in AI education and awareness


    Going into 2024, we should expect to see a surge in the implementation of both formal and informal digital skills and capacity-building initiatives. These initiatives are meant to increase digital and AI literacy, close the digital skills gap, develop new digital skills for the future, and upskill and reskill the African workforce. Senegal, in alignment with its AI Strategy, is committed to training 90,000 Senegalese in data science by 2028 (Republic of Senegal 2023). The country has a vision to be a leader in AI skills within the West Africa region and is building a sub-regional centre of excellence on data science and AI which will service West African countries. South Africa has launched a pilot programme in its schools to introduce new subjects for primary school learners which include AI and coding. These initiatives are still at a small scale and there is an urgent need to roll out wider programs targeting different societal groups. 

    Governments and private sector actors should also work together with academia and civil society to promote public awareness and AI literacy. To be more impactful, these AI programs should not be standalone, rather they should be integrated into data protection and cybersecurity and digital skills programs.

    Moving towards the bigger picture


    As AI policy discussions gain momentum on the African continent, there is a need for a comprehensive approach to AI governance. Policy interventions on AI should be integrated into broader initiatives addressing privacy, security, access to data, protection of intellectual property, human rights, and mechanisms for cross-border data sharing. A collective effort by diverse stakeholders is crucial to interconnect these policy nodes and weave them into a cohesive action plan for policymakers. I hope to see AU member states taking the initiative to implement the AU AI Strategy, the AU Data Policy Framework, the AU Convention on Cyber Security and Personal Data Protection, and the African Continental Free Trade Agreement with a special focus on mechanisms for data access, promotion of open data initiatives and data sharing across the continental free trade area. Good data governance remains crucial for the building of a successful AI ecosystem. These developments should take shape with the support of and involvement of diverse stakeholders including academia, civil society, the AI industry, multinational corporations, governments, data protection regulators, and human rights defenders. 

    Acknowledgements


    I would like to thank Chloe Teevan for her insights and feedback and Annette Powell for formatting. The views are those of the author and do not represent the position of ECDPM. All errors remain those of the author. Comments and feedback can be sent to Melody Musoni (mmu@ecdpm.org). 

    References


    AUDA-NEPAD. 2023. Artificial Intelligence is at the core of discussions in Rwanda as the AU High-Level Panel on Emerging Technologies convenes experts to draft the AU-AI Continental Strategy. 29 March.

    Bradford, A. 2019. The Brussels Effect: How the European Union Rules the World. New York: Oxford Academic.

    EC. 2021. Regulation of the European Parliament and of the Council laying down harmonised rules on Artificial Intelligence (Artificial Intelligence Act) and Amending certain Union Legislative Acts. 21 April. Brussels: European Commission.

    Egypt National Council for Artificial Intelligence. 2021. Artificial Intelligence Strategy.

    EU. N.d. European Union Global Gateway Strategy. Brussels: European Union.

    Erastus, E. 2021. Algorithmic apartheid? African lives matter in responsible AI discourse. Paradigm Initiative. 21 May.

    DigiWatch. 2023. Nigerian government calls on researchers to collaborate on national AI strategy. 29 August.

    Fana News. 2023. Ethiopia Finalizing National Policy On AI. Fana News. June 30.

    Government of India. 2023. G20 New Delhi Leaders’ Declaration. G20 Declaration. Ministry of External Affairs. 9 September.

    Government of Rwanda. 2023. The National AI Policy. Rwanda Ministry of ICT.

    Government of South Africa. 2020. 4IR Report - Summary Report & Recommendations of the Presidential Commission on the 4th Industrial Revolution. 23 October.

    Gwagwa, A, Kazim, E. and Hilliard, A. 2022. The role of the African value of Ubuntu in global AI inclusion discourse: A normative ethics perspective.

    Indeje, D. 2023. Keynote: AI in Kenya: Opportunities, Challenges and Policy Actions. KictaNet. 16 November.

    Mauritius Working Group on AI. 2018. Mauritius Artificial Intelligence Strategy.

    Musoni, M., Domingo, E. and Ogah, E. 2023. Digital ID systems in Africa: Challenges, risks and opportunities. ECDPM Discussion Paper 360. Maastricht: ECDPM.

    Musoni, M. 2023. Looking into the crystal ball: Artificial intelligence policy and regulation in Africa. ECDPM Commentary. Maastricht: ECDPM.

    Oxford Insights. 2023. Government AI Readiness Index. Authors: Hankins, E., Fuentes Nettel, P., Martinescu, L., Grau, G. and Rahim, S. (Ed. Rahim, S.). 20 December.

    Republic of Senegal. 2023. Senegal Artificial Intelligence Strategy.

    Terblanché, N. 2023. Namibia will focus on ethical AI development under UNESCO’s guidance. Windhoek Observer.

    UN Secretary-General. 2023. Secretary-General's Advisory Body Members - Artificial Intelligence. UN Statements 26 October.

    UNESCO and Government of Namibia. 2022. Windhoek Statement on Artificial Intelligence in Southern Africa Windhoek. 9 September.

    Vainilavičius, J. 2023. Deepfaked African Union chief called European leaders. Cybernews. 15 November.

    Loading Conversation