Multiple commentators warn that generative AI systems could reproduce or intensify inequality in South Africa if they are not designed, deployed, and governed carefully. As large language model–based chatbots and digital assistants become embedded in everyday digital services, access to these tools—and the benefits they provide—may not be evenly distributed. Sources emphasize that AI can reflect existing social and economic gaps through the data used to train models, the ways systems are targeted or made available, and how people can use them effectively. If groups with more resources have greater access to high-quality devices, connectivity, and digital skills, they may gain more from AI-enabled services, while others may be left with limited or poorer experiences. The guidance across sources is that inequality can re-emerge in digital form when safeguards are missing, including measures to ensure fair performance across different communities and accountability for harms. Overall, the discussion focuses on how AI’s widespread rollout could create new barriers or widen existing ones, rather than automatically reducing disparities.