Regulators around the world are in a race against time to come up with rules to govern the artificial intelligence (AI) space. The warnings that computers could soon have the level of human intelligence has raised a new regulatory headache for most governments. However, proponents warn that over-regulation could kill innovation and cut its benefits.

Innovations around AI have been moving so fast that regulators appear lost in formulating frameworks to govern the sector. The European Union (EU) is among the first to come up with a treaty, meant to make AI ethical and trustworthy. The treaty spells out global standards for responsible AI development and deployment and will influence other frameworks developed in different countries.

What will be the place of Africa in this fast-developing technology? At the beginning of 2024, there was a lot of talk about this being the year that most countries will pass AI regulations. Halfway through the year, nothing much has happened on the continent. Possibly this is why the African voice is missing in the drafting of the EU AI Treaty.

The treaty was drafted with input from countries outside the EU including the US, Argentina, Israel, Japan and Uruguay. While non-EU members are welcome to sign the legally binding agreement, the non-involvement of an African player could mean it does not contain local perspectives. All signatories to the treaty, which will be ratified on September 5, 2024, must ensure responsibility and accountability for the impacts that AI has on human rights.

The exclusion of Africa in the development of the treaty raises challenges for a continent burgeoning with innovation. Without Africa on the table, the new treaty risks worsening the data dependency that has plagued the continent in technological developments. Much of the current AI development relies on foreign technology and data housed outside the continent. With Africa not involved in the treaty process, it is unlikely that the companies involved in AI development would consider setting shop on the continent. This raises a serious concern, algorithms trained on non-African datasets may further existing racial biases, resulting in unfair outcomes.

The lack of diversity in datasets raises serious ethical challenges that could hamper the adoption of AI on the continent. AI has been touted as a technology that can be used to bridge existing societal gaps such as increasing financial inclusion among marginalised groups. But without proper safeguards, biased data used to develop AI algorithms could disproportionately affect African populations and other minority groups. For example, a surveillance algorithm developed without diverse data could inaccurately target certain communities.

Many African countries still lack comprehensive AI regulation. The Malabo Convention ratified by the African Union (AU) in 2023 guides the AI policy on the continent. This is not sufficient. Only Mauritius and Egypt have formulated AI laws while South Africa, Nigeria and Kenya have initiated plans to create regulatory frameworks. There lies a risk that the AI regulatory gap in Africa will widen if the continent does not sit at the same table with other countries. This exposes most African states to unregulated AI developments that could pose negative economic and social impacts.

Despite being excluded from the negotiations over the future of AI, Africa still has a role to play in shaping the treaty. The treaty is pushing for international cooperation, and with over 1.4 billion people, African states advocate for inclusive discussions to have local perspectives considered. By participating in global AI forums held by the United Nations (UN) and OECD’s Global Partnership on Artificial Intelligence (GPAI), the continent can reclaim the position it has been denied on the table.

African nations can collaborate and build a regional framework, picking lessons from the EU AI Treaty. This collaborative approach could promote a unified voice in the AI discourse and ensure that regulations are tailored to local needs. Regional economic blocs like the East African Community (EAC) or the Economic Community of West Africa (ECOWAS) can consider regional frameworks. For instance, it makes little sense for Kenya to develop an AI regulation, when EAC is advocating for further regional integrations including a monetary union.

The EU AI treaty represents the first step towards a global ethical AI standard. While there is a long way to go for Africa, there are some promising initiatives that show commitment to responsible AI development. Mauritius has developed a national AI policy that puts human beings at the centre. It emphasises fairness, transparency and accountability, ensuring the technology benefits everyone. Nigeria and Kenya are in the process of doing the same. The African Institute for Data Science (AIDS), a pan-African organisation, is promoting responsible data practices and has programmes to help African states build capacity for AI.

Adonijah Ndege, Senior Reporter – East Africa

Feel free to email adonijah[at]bigcabal.com, with your thoughts about this edition of NextWave. Or just click reply to share your thoughts and feedback.