In short

• AI risks deepening South Africa’s inequalities by displacing workers and widening digital divides.
• Power concentrates among those controlling data and infrastructure, reinforcing capitalist and global hierarchies.
• Ethical, culturally grounded, community-centred AI is essential to protect ubuntu, dignity and social cohesion.
• Inclusive governance, representation and community-based innovation hubs can ensure equitable AI benefits.

Artificial intelligence (AI) is transforming South Africa’s labour market, social systems and cultural identity. This article assesses its socio-economic effects from an African perspective, focusing on the voices of marginalised communities. It emphasises the need for ethical governance, inclusive innovation and policies that reflect local culture to ensure AI promotes shared values like ubuntu rather than worsening inequality or undermining human dignity.

Image by Susana Cipriano from Pixabay

Artificial intelligence (AI) refers to human-made systems and automated technologies that increasingly replace human labour. It forms part of a new global transformation that is reshaping traditional ways of working and living. In South Africa, this shift demands critical examination, particularly given the country’s deeply entrenched inequalities. South Africa has one of the highest inequality rates in the world, with a Gini coefficient of 0.63 and youth unemployment reaching 62.1% (based on the expanded definition of unemployment). While AI-driven technologies and innovations have the potential to be transformative, they also risk deepening social and economic divides.

In labour-intensive sectors such as mining, agriculture and manufacturing, automation is displacing workers and leaving already-vulnerable communities with fewer opportunities for generating livelihoods. Globally, AI often reinforces existing power hierarchies. Those who control data, digital infrastructure and capital tend to benefit the most, while others are excluded. In South Africa, this trend risks widening the digital divide and marginalising communities that lack access to technology or the skills to use it effectively.

AI as a reflection of capitalism At a seminar held on 26 September 2025, the authors presented their article, “Artificial Intelligence: A New Global System, a Genocidal Project, or the Revival of the Tokoloshe Culture?” The presentation examined how AI is transforming South Africa’s labour market and cultural landscape. The authors argued that without ethical and inclusive innovation, AI could deepen existing inequalities and erode African values such as ubuntu.

They noted that AI development and deployment are largely driven by capitalist imperatives, benefiting those with economic and technological power. As data becomes the new currency, those who control AI technologies and innovations accumulate disproportionate wealth and influence. This trend threatens to widen the gap between the Global North and South, as well as between the rich and poor within countries. In South Africa—where 10% of the population owns more than 80% of the wealth—AI could further entrench this imbalance.

Intersecting inequalities: AI, politics and corruption

The rise of AI intersects with South Africa’s existing political and economic inequalities. AI-driven automation in industries such as mining, agriculture and manufacturing—which employ many low-skilled workers—has led to job losses. Machine-operated drills, innovative farming tools and robotic assembly lines are efficient, but they leave behind communities that rely on employment as manual labourers. The result is economic displacement and a loss of dignity for those who are unable to transition into new roles.

This transformation is not only economic but also profoundly cultural. African philosophies such as ubuntu emphasise communal interconnectedness and shared humanity, while AI systems are often driven by individualism and profit maximisation. Replacing community-based workers in sectors such as healthcare or social services with automated systems may boost efficiency, but it risks weakening the relational bonds of trust and reciprocity that are central to African cultural life.

The seminar focused on a key question: What measures can be taken to ensure that AI innovations remain human-centred and ethical? Presenters and discussants explored strategies for aligning AI with human values and ethical principles, with emphasis on African contexts and social justice considerations.

Inclusive innovation and representation

Inclusive innovation and representation require the active involvement of diverse communities in AI design and implementation. Their perspectives and needs should inform how AI is developed and applied. Embracing local knowledge and cultural values can guide AI development towards promoting equitable access and collective well-being. Engaging ordinary citizens—not just experts or corporations—ensures that a broader range of insights and experiences shape decisions about AI.

Prioritising ethical development and cultural integrity

There is an urgent need to implement AI responsibly, considering social impacts alongside economic benefits. AI policy and governance must prioritise fairness, inclusivity, equity and social impact. While initiatives such as the SA Connect have helped to expand internet access in rural areas, more must be done to close the digital divide and prevent the exclusion of vulnerable populations. AI is not just a technological issue; it is also a social, cultural and ethical one. In South Africa, where inequality is deeply rooted, and cultural heritage is rich, the development of AI must be approached with care.

The article presented a provocative metaphor of AI as a “genocidal project”—not to suggest physical violence but to highlight the systemic erasure of livelihoods, cultures and human agency. Similarly, the phrase “data is the new oil” underscores data’s value and power in the digital age.

The authors refer to a “tokoloshe culture”, drawing on South African folklore to show how different communities perceive, understand and respond to new technologies. The tokoloshe—a mischievous and malevolent spirit believed to cause misfortune, illness or social disruption—serves as a metaphor for the anxieties surrounding AI. Recognising such cultural metaphors helps communities articulate their concerns and reimagine technology in more inclusive, human-centred ways. Such metaphors can be further expanded to encompass specific aspects of technology and innovation, their production and their use.

Dr Johannes Sibeko, coordinator of the Digital Humanities Hub at Nelson Mandela University, cautioned against attributing rising inequality solely to AI. He highlighted the role of deeper structural issues—such as governance failures and corruption—that exacerbate socio-economic inequality alongside technological change. The discussion emphasised the need for a holistic approach. Technological interventions, such as AI systems, must be accompanied by ethical governance and inclusive policy measures to ensure that marginalised communities are not further disadvantaged by either technological change or systemic political failures.

Towards inclusive AI adaptation in South Africa

South Africa has begun responding to these challenges through its National Artificial Intelligence Policy Framework, which aims to guide ethical AI development, promote inclusive growth and mitigate risks such as job displacement. However, successful implementation depends on whether South Africa can align technological progress with the country’s social and cultural values, and ensuring that marginalised groups share in its potential benefits.

The authors propose establishing Community-Based AI Innovation Hubs in under-resourced areas. These hubs would serve as collaborative spaces for communities, technologists and policymakers to co-create AI solutions that address specific social needs, such as improving healthcare access, offering local language translation or supporting agriculture.

Rooted in ubuntu and community values, such hubs would promote digital inclusion, foster skills development and ensure that AI systems reflect empathy, mutual care and collective well-being.

This approach would democratise access to AI while grounding its implementation in African epistemologies, ensuring that technological progress strengthens social cohesion rather than erodes it.

Research contacts and acknowledgements

This Review article was written by Zikhona Ngqula and Lebogang Khoza (PhD research trainees), with inputs from Dr Jacqueline Borel-Saladin and Dr Nazeem Mustapha from the HSRC’s Research, Development, Science and Innovation (RDSI) division. It is based on the paper “Artificial Intelligence: A New Global System, a Genocidal Project, or the Revival of the Tokoloshe Culture?” and the RDSI Internal Seminar held on 26 September 2025, where the paper was discussed with HSRC researchers and respondents. For more information, contact Ngqula at zngqula@hsrc.ac.za.

If you enjoyed reading this article, please click here to subscribe to the HSRC Review quarterly magazine.