/ 28 September 2023

Political and security considerations missing from Africa’s artificial intelligence strategy

Chatgpt2 Getty
OpenAI’s generative AI tools, ChatGPT and DALL-E 2, has brought artificial intelligence into contemporary popular discourse in all corners of the social sphere, from the art world and music to the impenetrable walls of the United Nations Security Council. (CFOTO/Future Publishing via Getty Images)

In 1956, researchers at the Dartmouth Summer Project mapped out the cyclical phenomenon of humanity’s social interest in artificial intelligence (AI). From this project, the terms “AI Winter” and “AI Spring” arose, the former denoting a period of cooled enthusiasm for artificial intelligence and the latter, a period of sustained and extensive interest in AI technologies. It is undeniable that we are living through an AI spring. OpenAI’s generative AI tools, ChatGPT and DALL-E 2, has brought artificial intelligence into contemporary popular discourse in all corners of the social sphere, from the art world and music to the impenetrable walls of the United Nations Security Council.  

Unlike other AI springs in history, such as that of the 1950s and 1980s, this renewed interest in artificial intelligence is geographically expansive and it has called for a serious governance initiative to mitigate the potential harms of AI, democratise its benefits and reign in reckless ambition. Many thinkers say that Africa simply lacks the political will and resources to harness and regulate artificial intelligence technologies, but is this true? What does Africa’s current AI governance agenda look like and what are some of the critical areas that have been overlooked? 

The roll-out of artificial intelligence and ancillary technologies in Africa in recent years has necessitated action. The growing prevalence of misinformation campaigns during elections and the predatory practices of artificial intelligence companies such as Sama and WorldCoin in countries like Kenya, Uganda, Sudan and Ghana have shown us just how necessary a regulatory governance framework is. Further inaction can only increase our continental vulnerability to cyber-colonisation, data predation and our relegation to a testbed for dual use technologies by tech-leading powers and corporations. 

Two key agencies and strategies map the continental artificial intelligence policy landscape, the African Union High Level Panel on Emerging Technologies (APET) and the panel’s 2021 report, AI For Africa: Artificial Intelligence for Africa’s Socio-Economic Development. 

The 10-member panel, which includes William Wasswa, Rachel Chikwamba and Francine Ntoumi and is chaired by Yaye Kène Gassama, acts as an advisory to the African Union. It has identified 10 emerging technologies that are priority areas for Africa; artificial intelligence is at number five. Convened by the AU’s development agency, it unsurprisingly prioritises Africa’s socio-economic orientation towards artificial intelligence. The panel explores how these technologies can be used for development and growth. The 2021 report by the AU High-Level Panel on Emerging Technologies also emphasises this orientation. Titled ‘AI for Africa: Artificial Intelligence for Africa’s Socio-Economic Development’, the 84-page document identified nine sectors in which AI was already being used on the continent: agriculture, mining, manufacturing, health, energy, education, finance, cultural heritage and public service delivery. 

The other key artificial intelligence strategy is earmarked for publication during the 36th Ordinary Session of the Assembly of Heads of State and Government of the African Union in 2024. Wasswa has also noted that the AU-AI Continental Strategy would be aimed at helping African countries develop AI-related tech products and services to assist with job creation in various sectors. It would also be aimed at improving data security.

As a continent that is primarily directed to undoing the economic disparity that has been a consequence of colonialism, heavy economic dependence on exports of primary resources and macroeconomic policy errors, the heavy socio-economic focus is not surprising. The thinking is that if artificial intelligence can truly propel African economies and help address the health, agriculture, climate and education problems then policymakers should seek to position themselves in a way that harnesses this growth. 

But the current ideation of artificial intelligence has left political and human security considerations wanting. Outside of the aforementioned political use of artificial intelligence in elections, the technology is also widely used in areas with high levels of violence to increase the coordination and capabilities of domestic security services and law enforcement. Fourteen countries, including South Africa and Kenya, use AI-driven surveillance and smart-policing platforms. Moreover, the creation and sharing of AI generated “deep fake pornography” targets women and gender diverse people at increasing rates

This pattern of usage has human rights, security and democratic implications. Misinformation during elections undermines democratic processes and political justice. Mass surveillance tools can strengthen domestic protection against criminality and violence but can also threaten human rights because it allows for targeted interventions against civilians. And the sharing of sexualised deep-fakes threatens progress made for women’s rights and freedoms on the continent. 

Although political and security considerations may seem far removed from the urgent reality of economic disparity, uncoupling them from policy and strategic interventions will not insulate Africa from their consequences. Moreover, our own aversion to foreign governance structures also means that Africa must develop its own security and political frameworks of artificial intelligence governance to prevent being steam rolled by supranational governance frameworks emerging from outside the continent. If continental institutions genuinely aim to provide African solutions to African problems, then a comprehensive strategy must be developed to will allow the continent to set its own technological agenda and mitigate artificial intelligence risks and harms as it works to harness its benefits. 

Suzie Shefeni is a technopolitics, policy and international affairs researcher based in Windhoek, Namibia, and an educational curator at the African Humanists for Ethical AI Collective.