(Designed on Canva)
Artificial intelligence (AI) is silently shaping the everyday lives of South Africans, often without their knowledge or explicit informed consent.
For instance, the South African Revenue Service (SARS) is using AI to enforce tax compliance, risk detection, and improve taxpayer interaction. The government deploys AI across departments such as the department of health, which uses AI chatbots such as MomConnect and Zuzi Chatbot to enhance public service delivery. Private organisations such as Tymebank use AI to improve their operations and customer experiences.
As these institutions adopt these technologies, algorithms increasingly make decisions that shape everyday life, often without citizens’ awareness or democratic input. Most South Africans do not know what AI is. In one particular survey, 73% of South Africans either have never heard of AI or have a minimal understanding of it.
When the public is unaware of how these technologies operate, algorithmic decisions that assist in governing a nation occur beyond democratic scrutiny. Democracy depends not only on voting, but on citizens being able to see, question, and influence the systems that govern them.
It was precisely this democratic gap that I addressed in a recent TEDx talk, The Hidden Danger of AI for Africa, which explored how AI systems imported from the Global North are often deployed in African societies without public participation, even though ordinary citizens bear the consequences of these technologies.
Take, for instance, the South African medical care system that uses AI tools to detect fraud and waste. The AI systems unfairly and falsely flagged black healthcare providers, resulting in unnecessary investigations, penalties, and experiences of racial discrimination.
The South African fintech and banking industry is deploying AI to automate decisions regarding loans. A study reported that AI created a 33-percentage point gap in approving loans to women-led construction businesses compared to men. This means women were 52% less likely to be approved for a loan, and when they did receive it, they got a lesser amount than men and at a significantly higher interest rate. This traps women in cycles of financial exclusion, preventing them from being able to compete fairly within the free market.
Moreover, in the South African law enforcement and private security sectors, AI is being adopted to prevent crime. AI tracks citizen behaviours, license plates, and assists in criminal profiling. This level of surveillance has been argued to erode the privacy and civil liberties of citizens, resulting in increased unneeded harassment of South Africans, especially in low-income communities.
Are South African citizens aware of these consequences presented by AI, and do they have any say in how it affects their lives?
Public opinion is not a “soft” element in developing the direction AI takes. Instead, citizen insight should be central to how these technologies are designed, deployed, and regulated. Currently, these conversations remain with academics, technologists, and the government.
Public pressure can be an essential mechanism in ensuring AI development is inclusive. Customers have historically boycotted private companies that use technologies that harm society. Public voters have the power to pressure governments that neglect policies that do not facilitate the safe and responsible development of AI in their communities. But in order for the public to play an active role in AI development, we must prioritise raising public awareness and understanding of these technologies.
To achieve this, the South African media must take on the responsibility for coverage and become a watchdog that continues to draw attention to how AI shapes society. The limited amount of journalism that reports on AI tends to frame or share stories of technological development and not stories of the consequences that these technologies have on society.
The Government and universities must promote AI literacy in order to set a common understanding of what AI is. They can do this by providing free online courses, as is currently being done by the University of Johannesburg. In Beijing, China, they have attempted to improve AI literacy amongst their population by making it compulsory for students as young as 6 to learn about AI, perhaps South Africa could learn from this initiative.
Academics working on an ‘AI for Africa’ initiative cannot limit these discussions to academic journals and conferences; they have a responsibility to disperse research in an accessible manner to the general public. The Humanities faculties, in particular, have a heavy responsibility to examine the societal impacts of these innovative technologies and help citizens engage with such issues to build an inclusive future.
AI is deciding who gets opportunities and who is left behind. If South African citizens cannot meaningfully participate in how these systems are deployed, then democracy itself is weakened. Public awareness should be the foundation for shaping an AI future that is inclusive for all and not just a privileged few.
James Maisiri is a doctoral student in the Department of Sociology at the University of Johannesburg.