The rapid pace of development in artificial intelligence (AI) has seemingly only been paralleled by calls for its regulation. Since there is broad consensus regarding the possible impact of unregulated AI, governments globally have responded through policymaking to mitigate this risk. Nevertheless, the overregulation of AI may, in turn, inhibit organizations from harnessing the advantages of AI and also limit a jurisdiction's ability to compete in this growing sector.

Accordingly, it is crucial that governments not only develop their AI sectors but also ensure effective regulation.

A number of African jurisdictions have taken steps to address this unique policy challenge, with various African countries implementing AI strategies, others still in the process of consultation with stakeholders on AI policies, and some African countries yet to make any pronouncements whatsoever on their approach towards AI regulation.

In this Insight article, Shahid Sulaiman, Davin Olën, and Monique Bezuidenhout, from Dentons, illustrate these three levels of AI regulation by unpacking the approaches of some key African jurisdictions, in the context of the draft AI policy of the African Union (AU), to AI regulation.

Degrees of adoption of AI regulation in African countries

The AU AI policy

The AU covers the vast majority of African states through its 55 members, having recently seen the re-joining of Morocco in 2017, following a 33-year absence. This significant membership of the AU lays the foundation for the realization of an Africa-centric path for the development and regulation of emerging AI technologies on the continent. The AU Development Agency published a draft policy (the AU Draft Policy) to advance this ideal on February 29, 2024, setting out a blueprint for AI regulation by member states.

The AU Draft Policy incorporates recommendations for industry-specific codes and practices, standards, and certification bodies. These recommendations are directed at assessing and benchmarking AI systems, providing regulatory oversight for the testing of AI, and establishing national AI councils. AU members without existing AI policies or regulations are encouraged by the AU Draft Policy to apply it as a framework to leapfrog the development of their national strategies. The AU Draft Policy also encourages AU members that already have AI regulations in place to review and align their policies with the AU Draft Policy to ensure a degree of consistency among AU members. Nevertheless, the official endorsement of the AU Draft Policy is only expected at the 2025 AU annual summit in Ethiopia.

Africa has already seen a progressive uptake of AI technology, investment, and, to a lesser degree, regulation. First among the African AI policymakers was Mauritius, followed shortly thereafter by Kenya and Egypt. Some African countries, like South Africa and Nigeria, are in the process of consulting with stakeholders about introducing AI policies, while others, like Morocco, are yet to propose any guidance on AI regulation.

Mauritius

In recognizing the potential of AI to address the country's financial and social issues, as well as its potential to be an important contributor to stimulate the country's economy, the Mauritian Government established the Mauritius Working Group (MWG) on AI. Following consultations coordinated by the MWG, the Mauritian AI Strategy was published in 2018.

Key considerations of the Mauritian AI Strategy include the potential applications and impact of AI, the establishment of an ecosystem to nurture AI development, and the creation of a regulatory framework for AI oversight. The Mauritian AI Strategy recognizes the ability of AI technologies to integrate into the broader Mauritian economy and support productivity, economic growth, and quality of life. For this reason, the MWG aims to have the Mauritian AI Strategy scaled up over the medium and long term to become a new pillar for sustainable growth and economic development in Mauritius.

Kenya

As with Mauritius, the Kenyan Government also recognized the economic possibilities of AI integration through the establishment of the Distributed Ledgers Technology and AI Task Force (Task Force) in 2018. In 2019, the Task Force published a report (the Task Force Report) noting the potential of AI to increase Kenya's global economic competitiveness and lay a foundation to position the country as a regional and international leader in Information and Communications Technology (ICT). The Task Force Report encourages increased investment in infrastructure and skills and further proposes strategies to balance innovation and regulation in the Kenyan private sector.

More recently, Kenya's Minister of ICT, Joe Mucheru, launched the National Digital Master Plan 2022-2032 (DMP). The DMP acknowledges the key considerations of the Task Force Report and echoes its encouragement for the uptake of research and development in the AI industry. However, in moving towards implementation, the DMP also sets out the Kenyan Department of ICT's strategy to deploy AI solutions. These solutions are aimed at attending to hurdles in Kenya's AI uptake and regulation as well as exporting such solutions to other countries.

This year has seen further steps towards a national AI strategy for Kenya through the collaboration of the Kenyan Government and the Deutsche Gesellschaft fϋr Internationale Zusammenarbeit (GIZ). This collaboration launched the 'FAIR Forward Artificial Intelligence for All Initiative' and is further supported by the German Federal Ministry for Economic Cooperation and Development and the EU.

In addition to national development, the DMP emphasizes the importance of developing international partnerships with leading actors in the AI industry, including those in the emerging technologies space.

Egypt

In a similar fashion to Kenya, Egypt emphasizes facilitating technology transfers and attracting foreign investments, which are discussed below.

The first phase aims to establish Egypt as a key role player in facilitating regional cooperation in AI. In comparison to Mauritius and Kenya, this phase of the Egyptian AI Strategy places more emphasis on the development of cooperative AI capabilities between Egyptian and foreign stakeholders, rather than its national or internal development. Harnessing AI's growth potential in Egypt only follows more substantively in subsequent phases of the Egyptian AI Strategy. Nevertheless, consultations regarding the second phase of the Egyptian AI Strategy are underway with Egypt's Minister of ICT and the NCAI.

The second phase of the Egyptian AI Strategy will be directed towards implementing various initiatives set out in the first phase and will take on priority initiatives aimed at enhancing investment, increasing public awareness, and developing greater AI capabilities in Egypt.

South Africa
Unlike Mauritius, Kenya, and Egypt, South Africa is yet to implement a national approach to AI regulation. Nevertheless, businesses in South Africa have seen an increased uptake of AI in their workplaces compared to the rest of Africa, with the result being a more pressing demand for the adoption of AI regulation. In addressing this demand, the South African Department of Communications and Digital Technology (DCDT) proposed a discussion document on AI (the Discussion Document) in April 2024.

The Minister of the DCDT's introduction to the Discussion Document highlights South Africa's ongoing development of an AI-focused legislative framework and outlines key Government AI priorities while setting policy development deadlines to ensure that South Africa is not left behind in the global AI race.

The aim of the Discussion Document is to facilitate interactions between the South African Government and the private sector. The interactions are, in turn, aimed towards establishing the appropriate principles to guide an approach towards AI regulation and develop a regulatory framework for AI. This approach varies from the African countries discussed above as it takes an extended consultative approach to policymaking, which includes wide engagement with the South African private sector and prolonged implementation.

In adopting this approach, South Africa has no existing legislation or regulatory framework to govern the use of AI. Thus, South Africa is heavily dependent on the application of data protection legislation for the regulation of AI. AI regulation in South Africa is presently enforced through the Protection of Personal Information Act (POPIA), the Consumer Protection Act, the Electronic Communications Act, and the Electronic Communications and Transactions Act, among others.
While existing legislation in South Africa has contributed somewhat to the regulation of AI-related developments, such legislation was not promulgated to serve the purpose of AI regulation.

South Africa is expanding its AI knowledge centers through the South African Centre for Artificial Intelligence Research (CAIR). CAIR's nine established and two emerging research groups span eight universities conducting AI research in South Africa. CAIR's funding is primarily provided by the South African Department of Science and Innovation and the country's AI research is further supported by the AI Institute of South Africa. The drafting of more specific AI regulations and the transformation of existing legislation to accommodate the changing and growing AI environment is becoming an evermore pressing concern for South Africa.

Nigeria
As with South Africa, Nigeria places reliance on existing legislation for the regulation of AI technologies, such as the Nigeria Data Protection Act (NDPA), among others. The NDPA places various obligations on organizations that perform activities related to the collection and processing of personal data but does not consider AI regulation specifically. Further obligations of the NDPA include taking responsibility for third parties involved in the processing of personal data, appointing a data protection officer (DPO), and implementing data processing security measures.

Nonetheless, Nigeria has moved towards implementing AI-specific policies. The Nigeria Federal Ministry of Communications and Digital Economy (FMCDE) directed Nigeria's National Information Technology Development Agency (NITDA) to develop a national AI policy for Nigeria. Following the FMCDE's directive, the NITDA published a white paper for the National Artificial Intelligence Policy (NAIP) in 2022. The NAIP is aimed at regulating the everyday implications of AI by addressing the complexities of AI economic integration. An updated draft of the NAIP was published in March 2023, with another white paper following closely thereafter in August 2023, this time announcing an expansion of the draft NAIP. More recently, the NITDA has hosted expert workshops on the NAIP to further coordinate a Nigerian strategy for AI.

Morocco

In a similar fashion to South Africa and Nigeria, Morocco relies on existing legislation for AI regulation. Morocco's existing laws include Law No. 09-08, which contains rules relating to personal data protection. Morocco has also established a National Commission for the Control of the Protection of Personal Data (CNDP). The CNDP acts as the regulator for data processing in Morocco.

While Morocco has progressed in its integration of AI into its economy through the establishment of academic institutions, large-scale data centers, and by hosting international conferences, no regulatory advancements have been made in the field of AI specifically. MoroccoAI, a private sector AI promotion initiative, took a step towards AI-specific regulation when it published recommendations for a Moroccan AI strategy in October 2023. Although this initiative indicates a step towards developing an AI-specific regulatory environment, no action has been taken by the Moroccan Government to develop and implement a national AI strategy as of yet.

These African countries showcase the three levels of progressive realization of AI regulation undertaken by some African jurisdictions.

Advancing AI regulation in Africa

It is clear from the above that a number of African countries are active participants in developing an equilibrium between AI regulation and innovation. It is particularly promising to identify jurisdictions focused on implementing AI strategies and policies for public sector reform, education, and research. Kenya, for example, has adjusted its national curriculum for schools in 2022 to introduce coding in primary and secondary schools' curricula. On the other side of the educational spectrum, South African universities and private associations host conferences and various other events in support of their development of local AI capacities.

The African AI innovation ecosystem is clearly active, but the progressive realization of AI regulation seems to be lagging behind in some jurisdictions. The immediate accompanying danger is that the risks associated with AI development and AI technologies may only be addressed subsequent to the realization of such risks. However, the AU Draft Policy, once adopted, has the potential to enable AU member states to leapfrog AI regulatory development, including those in the developmental and experimental stages of their AI regulation.

Originally published on Data Guidance

Authors: Shahid Sulaiman, Davin Olën and Monique Bezuidenhout