AI in Africa: 4Sight aims to move artificial intelligence beyond chatbots

Rands 1.16 billion in revenue and a 45.8% increase in operating profit. South African technology group 4Sight believes the next phase of AI will no longer be limited to ChatGPT, Copilot or conversational assistants. Companies must now rethink their processes around systems capable of analyzing, deciding and acting.

AI in Africa: chatbots are no longer enough

South Africa already has a head start in the use of generative AI. 23.1% of its working population was using a generative AI tool at the beginning of 2026. For 4Sight, however, this initial adoption phase is only the beginning.

The company wants to move organizations from individual prompts and copilots to systems integrated into their operations.

An AI agent could, for example, receive an invoice, check the information, consult an ERP system, trigger an approval request and forward only unusual cases to a human. AI is no longer simply being asked to write a summary. It is intervening in the process itself.

4Sight describes this approach as “Automated Intelligence.” Its platform combines AI, robotic process automation, low-code tools, data and business software.

On its automation infrastructure, the group presents agents capable of orchestrating workflows across documents, emails, ERP systems and CRMs.

The results are beginning to emerge

The argument would be less compelling without the figures.

For the financial year ended in February 2026, 4Sight reported R1.163 billion in revenue, up 16.3%. Its operating profit jumped 45.8% to R71.7 million. Adjusted earnings per share increased by 46.1%.

The group attributes part of its productivity gains to the internal integration of AI.

This shift reflects a broader trend: Africa is beginning to build more local AI infrastructure and capabilities, moving beyond simply consuming American models.

At 4Sight, four elements must advance together: data, processes, training and change management. A company can buy Copilot tomorrow morning; if its data is fragmented and its processes poorly designed, AI will mainly automate the existing disorder.

The group also strengthened its leadership team this week to accelerate its “AI-first” strategy.

African AI moves into operations

4Sight is now targeting broader expansion across the continent. The group had already reported triple-digit growth in markets such as Ethiopia, Liberia, Sierra Leone, Namibia, Uganda, Ghana, Kenya and Zambia.

Finance, mining, industry, human resources and public services: use cases are becoming far less spectacular than a chatbot capable of speaking like a human. They may also be far more profitable.

The shift obviously raises new questions. An agent that drafts an email can make a mistake. An agent authorized to modify an order, process an invoice or act within a financial system can be much more costly when it makes a mistake.

Governance, action logging, permissions and human validation are therefore becoming essential. This issue will quickly enter regulatory discussions, as COMESA is already preparing a common AI strategy for its 21 member states.

AI in Africa is therefore entering a less visible and probably more important phase. After learning to converse with machines, companies are beginning to entrust them with part of their work.