1. Core Claim
AI is not just another IT tool. It will re-organise capability and power in society. South Africa must decide which layers of the technology to own, or else risk value leakages to multinationals abroad.
2. Why It Matters
- Pace of AI outstrips policy and bureaucracy responses, creating a governance debt.
- Current deployment is dominated by a few global actors, without consent nor compensation (appropriation without compensation).
- The Global South has historically under-captured value from technology disruptions/waves, e.g., electricity, internet, and mobile.
3. Opportunities
- Leapfrog potential in solving “wicked problems” of society: health, education, climate, public safety.
- Building contextual and sovereign AI systems that are rooted in African/South Africa realities.
4. Layers South Africa Must Own
- Data & local context (languages, norms, routines, culture, processes).
- Evaluation & assurance (local benchmarks, assurance, bias testing).
- Domain ontologies & workflows (how work is actually done.
- Change management (skills, roles, incentives, social nets).
5. Risks Inherency
- Bias & inequality.
- Job disruption without task redesign.
- Concentration of power via global platforms.
6. Takeaways
- AI is intelligence; intelligence is leverage.
- Governance is execution, not speeches or noble intentions.
- Hybrid systems with a strong sovereign niche are the pragmatic roadmap.
7. Closing Thought
If Africa does not choose its participation layers, others will choose for us. The task is to bend this wave toward our priorities in a responsible and ethical manner.
Download a copy of the presentation: AI in the Global South: A South African Perspective