The use of AI agents by large companies in South Africa is still low compared to other countries, a fraud prevention company said at a roundtable on AI in banking and payments.
The event was organised by Sumsub, a company that offers identity checks, fraud prevention, transaction monitoring and risk management tools. IOL reported on the discussion.
“Major enterprise adoption of true AI agents in South Africa is still quite low compared to global markets,” said Jarryd Jensen, Sumsub’s Regional Director for Southern Africa.
“South Africa boasts a sophisticated market with advanced digital banking and instant payment systems, and consumers are already very comfortable with digital onboarding. However, full agentic adoption is still in its early stages,” Jensen said.
Hannes Bezuidenhout, Sumsub’s VP of Sales for Africa, said some banks are moving ahead. He named Discovery Bank as one bank pushing into AI agents.
Bezuidenhout said South Africa as a whole still lags behind leading countries such as China.
Agentic AI means letting AI tools make bigger decisions for a person. Examples include managing investments, running business tasks with many steps, or booking a full trip.
The roundtable was led by Mick Amelishko, an AI Advocate at Sumsub. He said the main change is that AI can now act on its own.
“Last year, AI was just an assistant. It helped check writing or offered basic suggestions,” Amelishko said.
“Now, AI can access the external world, read the live internet, interact with other digital products, make independent purchases, register on websites, or interact directly with bank accounts if granted permission,” he said.
Amelishko said society has worked for centuries on the idea that every transaction was carried out by a person. He said that human part is now slowly being replaced.
Speakers at the event also discussed who is to blame when an AI agent makes a mistake. The question is whether the person who gave the instruction or the machine that acted should be held responsible.
The general view at the roundtable was that AI itself cannot be legally blamed for unlawful acts. Speakers said responsibility must always lead back to a human operator or the person in control of the system.
According to IOL, new kinds of financial fraud are growing fast, and rules often fall behind new technology.
Data from Sumsub shows the global identity fraud rate dropped slightly to 2.2% in 2025. It was 2.6% in 2024.
Over the same period, cases of highly sophisticated fraud rose by 180% year on year, according to Sumsub. The company said criminals moved away from simple scams towards advanced attacks that use AI.
Source: this article is based on reporting by IOL. Image: IOL. Written with the help of AI and published by the Tzaneen Voice Technology desk. See our Editorial Standards.
Werner Jacobs covers technology for Tzaneen Voice. Werner reports on mobile networks, data prices, internet access, smartphones, apps, AI, cybersecurity and South African tech companies. Each story explains what a new product, price change or policy means for ordinary users in South Africa.