Capitec is investing heavily in its AI capabilities as part of a push to stand out in an increasingly digitalised world of payments and personal finance.
AI investment has grown exponentially in recent years, sparked by the rapid adoption and popularity of OpenAI’s ChatGPT since it was launched in November 2022.
The country’s largest retail bank, with 26-million customers, said investment in systems and platforms continued to drive the increase in its IT expenses for the year ended February 2025.
Excluding salaries, a major factor in technology spend across South Africa’s banking, IT expenses grew by 12%. Including staff costs, IT expenses rose 17.5% to just more than R3bn, from R2.559bn in the year before.
Cloud-based computing fees increased 25% “as the utilisation of cloud data capabilities grew and allowed us to scale our operations further”, the bank said.
This hike in cloud computing is likely to be tied to growing use of AI and the development of systems by the bank.
“We are building a responsible AI framework based on a commitment to unlock value for our clients,” Capitec said in its annual report for the period.
The bank says AI is being used in real-time fraud prevention, risk modelling and upskilling employees.
The bank also touted its recently developed Capitec Pulse AI that “provides real-time, contextualised client information to client support agents, and we plan to use AI in the personalisation of client experiences at scale and to further enhance financial inclusion”.
“Our investment in AI for the year was significant.”
The bank’s integration of AI agents hints at an organisation willing to embrace the latest technologies if it means staying ahead.
This is part of a growing trend known as “agentic AI”, in which systems take over more tasks from human beings. For example, an agent can function like a personal assistant, making bookings, scheduling meetings, and summarising notes and other important information.
For a bank, this includes AI agents that can handle customer queries, give financial advice, or even approve a credit or loan application.
Over the last year, South Africa’s established lenders have expressed scepticism about giving AI agents such autonomy. The government’s recent withdrawal of its draft national policy on AI, after discovering fake research, serves as a proof point of the risks that may come if the technology goes unchecked.
Even then, Capitec said, “AI is not a future aspiration — it is already at work. We invested significantly in AI this year.”
The bank said AI-driven fraud models “protected” clients from R673m “in potential losses” during the period, adding that generative AI is live in its compliance operations.
“An AI agent is embedded in business banking credit processing, with a deliberate strategy to scale it across the division. And our people are using these tools — not occasionally but daily.”
The bank said almost 5,000 employees hold active AI licences, with usage per employee averaging four times a day.
This is not to say Capitec does not understand the risks of AI.
The Cape Town-based lender highlighted that technology, cybersecurity and financial crime risks “remain inherently elevated given an increasingly sophisticated and interconnected threat landscape, while rapid advancements in AI introduce additional complexity”.
“The expanding use of advanced analytics, machine learning and emerging AI capabilities increases exposure to model risk, data dependency, explainability challenges and evolving regulatory expectations.
“These risks are being actively managed through strengthened model governance, responsible AI frameworks, enhanced oversight and continued investment in preventative, detective and response controls.”
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.