Seen Here: President Ntuli, Managing Director of HPE South Africa, says continuously available, AI-powered fraud detection is becoming essential as South African banks confront increasingly sophisticated digital financial crime. He believes agentic AI will play a pivotal role in helping financial institutions strengthen security, reduce fraud losses and protect customer trust in an evolving digital banking landscape. Photo Credit: Supplied
JOHANNESBURG – As digital banking fraud escalates across South Africa, financial institutions are under growing pressure to strengthen their defences against increasingly sophisticated cybercriminals. Industry experts now believe that continuously available, AI-driven fraud detection systems are no longer a competitive advantage but a business and regulatory necessity.
According to President Ntuli, Managing Director of HPE South Africa, the rapid evolution of financial crime is exposing the limitations of traditional fraud prevention systems, which were designed for a far less complex digital landscape.
“Trying to stop modern fraud with traditional tools is a bit like guarding a busy highway with a clipboard and a whistle,” Ntuli said.
“The rules are clear, but the traffic has changed. It is faster, more complex and increasingly automated. By the time something looks suspicious, it has already sped past.”
Digital fraud reaches alarming levels
Commercial crime continues to rise sharply in South Africa, with reported cases nearly doubling over the past decade.
The online banking environment has become particularly vulnerable.
Digital banking fraud incidents surged by 86% during 2024, with financial losses exceeding R1.8 billion, highlighting the growing sophistication of cybercriminals targeting banks and consumers.
Unlike traditional criminal activity, today’s fraud is continuous, coordinated and designed to exploit vulnerabilities in real time.
Criminal syndicates increasingly monitor banking systems for planned maintenance, network outages or cybersecurity disruptions—moments when monitoring capabilities may be reduced and fraudulent transactions are more likely to go undetected.
Adding to the challenge, more than half of banking executives worldwide believe organised criminal networks are evolving faster than financial institutions can respond.
Why agentic AI is changing fraud prevention
To combat this growing threat, financial institutions are increasingly turning to agentic artificial intelligence (AI)—a new generation of autonomous AI systems capable of analysing, learning and responding to suspicious activity almost instantly.
Unlike traditional rules-based systems that rely heavily on manual intervention, agentic AI continuously monitors transactions, evaluates risk and initiates responses within seconds.
According to Ntuli, this technology is transforming fraud prevention in five key ways.
Instant response to suspicious activity
Traditional fraud investigations often depend on human analysts reviewing alerts before action can be taken.
Agentic AI dramatically reduces that delay.
The technology can identify suspicious transactions, analyse supporting evidence, assess potential risk and automatically trigger security measures—such as additional authentication—within less than a second.
This rapid response significantly improves the chances of stopping fraud before financial losses occur.
Reducing financial losses
Real-time fraud detection becomes especially valuable during peak transaction periods when criminals are most active.
International research indicates that fraud attempts on credit card networks can increase between two and nine times during high-volume shopping periods.
AI-powered monitoring systems that continuously scale with transaction volumes enable banks to maintain performance while preventing fraudulent activity, even during periods of exceptional demand.
Empowering human fraud investigators
Rather than replacing fraud analysts, agentic AI is designed to enhance their productivity.
Routine fraud attempts can be blocked automatically, allowing human investigators to focus on more sophisticated criminal schemes.
Instead of receiving basic alerts, investigators are presented with comprehensive case files that include transaction histories, customer profiles, behavioural analysis and supporting evidence, enabling faster and more informed decision-making.
Fewer false alarms
One of the long-standing weaknesses of conventional fraud detection systems is their inability to understand context.
A traditional system might flag a South African customer’s purchase in Thailand as suspicious simply because it occurs overseas.
An AI-powered system, however, can recognise that the customer previously booked flights and accommodation in Thailand and determine that the transaction is legitimate.
Reducing these false positives improves customer experience while allowing banks to focus on genuine fraud risks.
Detecting sophisticated criminal networks
As cybercriminals increasingly use artificial intelligence to orchestrate attacks across multiple payment platforms and authentication systems, detecting complex fraud schemes has become significantly more challenging.
Agentic AI is capable of analysing activity across multiple systems simultaneously while incorporating external threat intelligence to identify emerging attack patterns.
Combined with highly resilient infrastructure featuring zero-trust security, continuous availability and advanced threat detection, these systems make it considerably more difficult for criminals to exploit weaknesses within banking networks.
Compliance becoming as important as protection
Beyond preventing financial losses, banks now face increasing regulatory expectations regarding the responsible use of artificial intelligence.
In November 2025, South Africa’s Financial Sector Conduct Authority (FSCA) and the Prudential Authority (PA) released a landmark report on artificial intelligence in the financial sector.
While the report does not introduce new legislation, it outlines clear expectations for how financial institutions should govern, monitor and manage AI technologies.
Industry experts believe the guidance effectively establishes a new operational benchmark, making continuous and reliable fraud monitoring an important component of regulatory compliance.
Trust remains the banking industry’s greatest asset
Ntuli believes the consequences of failing to modernise fraud prevention extend well beyond financial losses.
Fraud not only erodes revenue but also undermines one of the banking sector’s most valuable assets—customer trust.
Consumers increasingly expect financial institutions to protect their money through intelligent, always-available security systems capable of responding instantly to emerging threats.
As digital payments continue to accelerate and cybercriminals become more technologically advanced, South Africa’s banking sector faces a pivotal challenge.
The future of fraud prevention will depend not only on adopting artificial intelligence but on deploying resilient, continuously available infrastructure capable of protecting customers, ensuring regulatory compliance and safeguarding institutional reputations in an increasingly connected digital economy.
