For many South Africans, few phone calls are as inconvenient as one-to-one calls debt collector.
Whether a payment is missed due to financial hardship or simply ignored, the conversation can often feel personal, stressful, and extremely embarrassing.
But as the country faces increasing financial pressures, a surprising change is beginning to emerge.
progressively, some consumer seem more comfortable discussing their overdue accounts with artificial intelligence Compared to a human agent.
According to Bruce von Maltitz, chief executive of customer engagement technology company 1Stream, AI voice agents are helping to remove some of the stigma traditionally associated with debt collection.
Von Maltitz said, “Debt is a difficult topic under the best of circumstances. When the debt becomes overdue or the subject moves toward debt collection, things become even more difficult, practically and emotionally.”
The household debt burden in South Africa continues to increase.
AT20's latest Credit Stress Report shows that nearly 40% of the country's credit active population is more than three months behind in at least one loan repayment.
Additionally, recent TransUnion research indicates that approximately 35% of South Africans expect they may miss at least one bill payment during the current payment cycle.
For businesses, these numbers represent a growing challenge, particularly in the retail sector, where Experian recently reported a default rate of 15.77% on retail credit products. There are currently around 22.2 million retail loan accounts across the country.
Von Maltitz believes AI can help both consumers and businesses navigate this difficult landscape more effectively.
“There are people who have serious financial problems that are causing them to default, but there are also many cases where the debit order did not go through because of a mistake, or a payment was missed and the customer fully intended to pay,” he said.
“In those cases, you often need a fast, consistent way to inform the person and help resolve the problem.”
Historically, collecting on small retail debts has often been uneconomic for businesses, with the cost of maintaining large collection teams exceeding the value of many overdue accounts.
Von Maltitz said, “The value of retail credit accounts is comparatively small. But with so many accounts active, recovery on overdue debt can quickly become meaningful to the business.”
However, AI voice agents are changing that equation by providing a scalable and cost-effective solution that can contact thousands of customers instantly, while providing what some consumers view as less judgmental interactions.
“AI voice bots have no ego, no mood, and no ability to make personal decisions, so consumers feel much less embarrassment,” von Maltitz explained.
“It simplifies the conversation. The bot is only there to confirm details and arrange the payment path.”
He emphasized that artificial intelligence should not completely replace human agents, but should instead be used where it provides the greatest value.
“That's the first interaction where we see demand. If you can get a promise to pay using a bot, it's quick and cheap,” he said.
“If the situation is more difficult or complex, you hand the call over to a human agent who remains essential in cases where there is conflict, crisis and the need for discretion and empathy.”
Research into AI-powered collections remains relatively new, but studies into broader AI and human interactions show that people are often more willing to engage with chatbots when discussing sensitive or potentially embarrassing topics, while still preferring human assistance for emotionally charged situations.
Apart from reducing inconvenience, AI can also improve debt collection results by enabling businesses to connect with customers sooner.
“A company that reaches out to the customer at the beginning of the month, while money is still available, has a better chance of resolving the account,” von Maltitz said.
“AI can help businesses make contacts quickly and at scale instead of slowly working through a list based on how many agents they have available.”
However, he cautions that these technologies must be deployed responsibly.
“This must be done carefully as consumers are already overwhelmed with spam calls, so AI collection needs to be part of a thoughtful communications journey, supported by SMS, email and clear identification of who is calling and why.”
As financial pressures on South African households continue to grow, increased AI-driven customer engagement could provide unexpected benefits. For businesses, it provides a sustainable way to recover low value debts. For consumers, this may offer a less intimidating first step toward resolving financial difficulties before they turn into something more serious.