In the second part of our feature on AI and the channel opportunity, we asked our panel what the main AI use cases are today, and where they expect AI adoption to scale to long-term.
Ebrahim Essop, senior automation solution manager at Nedbank CIA Robotics
Currently, AI is being used in software engineering, marketing, customer service automation, document processing, fraud detection and decision support. Over the long term, we can expect deeper integration into core business processes, including personalised customer experiences and intelligent decisioning across the enterprise.
The way we work has changed and will continue to evolve, we have gone from spending hours upfront on analysing problems and designing solutions and are evolving into masters of prompting. Specialists with AI are ideally equipped to amplify their skills by giving critical context to the AI and then reviewing the output and adjusting course as needed. Imagine having a builder that can create just about anything digital with the only limitation being your imagination and token credits.
Werner Herbst, MD of First Distribution
The current use cases are highly practical and value-driven. Organisations are focusing on improving productivity, automating manual processes, and enhancing customer engagement.
We’re also seeing growing interest in AI agents — not just tools that assist individuals, but systems that can execute tasks and workflows more autonomously.
In Africa, AI is increasingly being applied in sectors like financial services, agriculture, and healthcare, where it helps address real capacity and scalability challenges.
Long term, AI — particularly agent-based models — will become embedded in how businesses operate.
We will move from people using tools, to AI orchestrating end-to-end workflows across systems. This will fundamentally change operating models, with human and digital workforces working together.
The real value will come when organisations redesign processes around AI, rather than simply layering it onto existing ways of working.
Ryan Martyn, co-founder and chief marketing officer of Syntech Distribution
Right now it’s mostly individuals within organisations using generative AI informally, for drafting, research, summarising and general productivity, rather than AI being embedded into core business processes. Few companies have moved past this experimentation phase into structured, measurable use cases with defined ROI.
Agentic AI is going to be the big shift, particularly in optimising workflows. But this only works if corporates are disciplined about it. They need to be clear on the specific areas where they want to introduce AI, and set clearly defined expectations for ROI and output upfront, rather than deploying it broadly and hoping for results.
Othelo Vieira, technical product manager lead at Cloud On Demand
The dominant use cases right now are where the ROI is fastest and most measurable. In financial services: fraud detection, credit risk, and automated customer service. In retail and telecoms: personalisation, churn prediction, and network optimisation. In the public sector and healthcare, there’s growing interest in AI-assisted diagnostics and service delivery — though regulatory caution slows rollout. On the enterprise side, document processing, contract analysis, and HR automation are gaining traction. Generative AI has also opened up a new wave of use cases around internal knowledge management, code generation, and marketing content — these are the fastest-growing category right now.
Long-term, AI becomes infrastructure — embedded in virtually every business process rather than deployed as a standalone capability.
I expect the biggest shifts in agriculture (precision farming and crop analytics, which Africa has a genuine global opportunity in), logistics and supply chain, education (personalised learning at scale), and healthcare diagnostics.
At the national level, AI-enabled e-government services will become a major battleground — countries that get this right will have a material advantage in service delivery and economic competitiveness.
The other big long-term story is AI agents: autonomous systems that don’t just assist humans but complete multi-step tasks independently. That transition is already beginning.
Barry Buck
Right now it’s content understanding and document intelligence — the work traditional automation could never touch. Reading the intent behind an email; pulling clean data out of a messy scanned document that would defeat an OCR system; then acting on it. Those are today’s big wins.
But the real prize, still largely untapped, is automating how we work — a team’s daily rhythm, not just its paperwork.
Andre Hogewoning, chief operating officer at Business AI
The majority of companies begin their AI journey in a practical and immediately valuable place — deploying domain-specific AI assistants designed to unlock the vast institutional knowledge that has accumulated within their organisations over years, often decades.
However, the use case itself is not the primary consideration when beginning an AI deployment. What matters far more is the framework within which that use case will operate.
Organisations must address critical dimensions before deployment: security and AI hazard management; compliance, non-binding use of LLMs’ model selection, token economic managementl and AI infrastructure.
Companies also need to urgently close the corporate IP gap, ensuring there’s no unintentional exposure of corporate IP.
Thereafter, secure AI platforms can enable the rapid deployment of multiple use cases without compromising security or compliance.
One of the most significant developments anticipated in the near term is the rapid mainstream adoption of AI agents — autonomous AI systems capable of executing complex, multi-step tasks with minimal human intervention.
The commercial implications are profound. AI agents are already being deployed globally to perform functions traditionally outsourced to specialist service providers — including marketing, content creation, research, data analysis and customer engagement — at a fraction of the conventional cost.
Perhaps the most commercially disruptive trend identified is the growing challenge to traditional enterprise software — specifically large-scale ERP (Enterprise Resource Planning) systems that have long been the backbone of major organisations worldwide.
At the individual level, it is anticipated that a transformation in how professionals manage their working day will follow shortly. The emergence of personal AI assistants — capable of handling scheduling, correspondence, research, reporting and administrative coordination — means that every employee will effectively have access to a small team of intelligent assistants working on their behalf.
Ziaad Suleman, senior vice-president and CEO at Cassava Technologies, SA and Botswana
Based on our client engagements across South African industries, four AI use-case clusters now dominate in production, not pilots.
In financial services, the flagship applications are fraud and financial-crime detection (AI models routinely preventing hundreds of millions of rand in client losses annually) and generative AI productivity and customer service assistants rolled out to entire workforces.
In the public sector, machine-learning risk-profiling has become the single highest-value use case, credited by revenue authorities with adding hundreds of billions of rand to collections through refund-fraud prevention and compliance targeting.
In retail, consumer and healthcare, hyper-personalisation and demand intelligence lead, with machine learning for delivery-route and demand optimisation, personalised loyalty and shopping assistants, and predictive AI health-guidance platforms serving millions of members.
And in mining, energy, manufacturing and agriculture, the pattern is operational and social-impact AI: predictive maintenance, computer-vision safety, autonomous survey and quality-defect systems, AI-augmented medical screening, and a growing wave of homegrown sovereign capabilities, such as African-language models and precision-agriculture analytics.
Long-term, AI will evolve from basic operational tools into intelligent predictive systems that optimise core industrial sectors and public governance. We anticipate the technology becoming deeply embedded in healthcare, national research, and infrastructure management, run entirely on localised sovereign networks.
Ravi Bhat, commercial solutions and AI officer at Microsoft South Africa
Today’s leading AI use cases are practical: employee productivity, customer service, software development, cybersecurity, finance and IT operations. The Microsoft 2026 Work Trend Index shows 49% of Microsoft 365 Copilot interactions support cognitive work, signalling that AI is increasingly being used as a thought partner, not only a writing assistant.
In South Africa, the strongest early adoption is in cybersecurity, administration, finance, IT infrastructure and customer service. The long-term opportunity is bigger: moving from tool rollout to workflow redesign through custom copilots, AI agents, industry solutions and multilingual AI embedded into core processes – the next phase of AI is not more pilots – it is redesigning how work gets done.