South Africa’s major banks are entering a new phase in their artificial intelligence (AI) journey in 2026, shifting from pilot projects and isolated experiments, to large-scale deployment across core operations.
After years of testing machine learning and data-driven tools, the banks are now accelerating investment and integration of AI into fraud detection, customer service, risk management and digital platforms, signalling the technology is becoming central to competitiveness in the financial sector.
In e-mail interviews with ITWeb, the local banks were in unison that they will ramp up the deployment of the technology in 2026. The institutions also revealed they will be pumping more money into technologies such as cloud computing, while beefing up their cyber security systems.
Johan Maree, CEO of FNB Nav, says in line with global trends, the South African banking sector is expected to shift towards large-scale AI deployment, moving beyond experimentation to industrial execution.
He notes the Fintech Association of South Africa indicates that approximately 45% of South African banks have each allocated more than R30 million for AI investments this year, with a strong focus on operationalising autonomous agents and predictive security capabilities.
“This shift marks a fundamental change in how financial services will be delivered, secured and regulated across the country,” says Maree.
Core banking transformation
Jörg Fischer, group chief information officer (CIO) of Standard Bank, believes AI is expected to play a central role in reshaping SA’s banking landscape through enhanced customer experience, improved risk management and greater operational efficiency.
“Standard Bank has publicly stated that AI is already transforming core banking functions, ranging from risk management, fraud detection, personalised client experiences and operational efficiency, and will continue to deepen its impact across the sector,” says Fischer.
He adds that the bank emphasises a “human in the loop” model, ensuring AI augments rather than replaces human judgement.
“This approach supports responsible, ethical and trust-centred AI deployment, an increasingly important factor as more financial services are automated. Additionally, Standard Bank’s digital transformation initiatives – driven by AI, cloud computing and automation – have already enabled an increase in digital transactions, signalling the acceleration of AI’s role in shaping consumer behaviour and digital adoption in South Africa.”
For Robert Benvenuti, CIO of data and applied AI at Absa, AI is a major technological advancement which will be as impactful to every aspect of banking as the introduction of the personal computer.
“2026 will be a year when banks move beyond just experimenting with AI, to use enhancing and optimising core processes, not to replace the human, but to make the human significantly more productive. AI won’t replace people, but people who use AI will replace people who don’t use AI,” says Benvenuti.
From Nedbank, Portia Matsena, divisional executive for hyper automation, says AI is moving decisively from experimentation to scaled, value-driven deployment across the banking industry.
According to Matsena, the focus is increasingly on embedding AI into core processes to improve decision-making, enhance customer experiences, and strengthen risk, compliance and operational resilience.
“A particularly significant area of impact is credit decisioning and credit management, where AI is enhancing the quality, speed and consistency of credit assessments while remaining grounded in strong governance and regulatory compliance. Rather than replacing traditional credit models, AI is augmenting them − enabling better risk differentiation, more informed lending decisions and more sustainable credit outcomes,” she says.
Andrew Baker, chief technology officer at Capitec, points out that AI will be a major driver of both innovation and risk in South African banking this year.
“We’re seeing continued expansion in the malicious use of AI through fraud scams, deepfakes and highly-sophisticated malware. Capitec is using AI to counter these risks by detecting malware, identifying suspicious or irregular payments, and proactively intercepting and delaying it.”
Agentic AI push
As South African banks accelerate the deployment of AI, their focus is increasingly shifting towards agentic AI – systems capable of acting autonomously, making decisions and executing tasks with minimal human intervention.
Unlike traditional AI tools that primarily analyse data or support predefined workflows, agentic AI is being positioned to transform core banking processes, from real-time fraud prevention and credit assessment to personalised customer engagement and operational automation.
Baker says AI is accelerating how banks solve client problems and deliver improvements. “At Capitec, we’ve focused on using AI to remove friction and resolve issues faster. We’ve recently launched Capitec Pulse, an agentic solution that analyses a client’s recent activity while their call is inbound to our call centre.
“In around 10 seconds, it scans real-time logs, error logs, databases and audit trails, then suggests the likely reason for the call. This reduces a common frustration for clients who would otherwise need to explain problems like being unable to log into the app, a failed payment or hitting a limit. Previously, this level of analysis could take days,” Baker explains.
He adds that Capitec has also completely re-engineered its fraud and cyber stack. “Last year, this enabled us to significantly grow payment volumes while reducing fraud by around 20%. As we invest further in AI-driven fraud solutions, we’re seeing a material reduction in fraud losses, including in areas that are notoriously difficult to defend against, such as authorised push payment fraud and invoice fraud.
“Beyond AI, our technology investment remains anchored in fraud prevention controls and frameworks, and completing the re-engineering of our business bank, with these projects being developed with our preferred cloud provider, Amazon Web Services.”
Nedbank’s Matsena is of the view that agentic AI represents a shift from task-based automation, to more goal-oriented, end-to-end process optimisation.
“In practice, this is improving turnaround times, reducing operational friction and enabling earlier interventions across areas such as credit monitoring and collections. From a credit management perspective, AI-enabled early warning systems and portfolio analytics are allowing banks to move from reactive to proactive risk management − supporting customers earlier while protecting balance sheet health. Investment in this space continues, with a strong emphasis on human oversight, explainability, auditability and control,” she says.
Maree from FNB explains that banks are replacing static digital interfaces with natural language, semi-autonomous agents capable of completing end-to-end tasks, from loan adjudication to dispute resolution.
“These systems are also becoming more anticipatory through life event modelling, allowing banks to offer advice and products as customers need them. In a diverse market like South Africa, multilingual AI in languages such as isiZulu and isiXhosa is also driving deeper financial inclusion.”
From trend to tangible
Standard Bank has identified the rise of agentic AI as one of the most significant emerging trends shaping the financial services sector.
According to Fisher, the technology represents a major opportunity for innovation and monetisation, echoing earlier views expressed by group chief economist Goolam Ballim, who highlighted agentic AI as a key driver of growth from 2025 onwards.
He says the bank is already deploying AI-powered tools across its African operations, spanning customer engagement analytics, AI-enabled customer relationship management platforms, agentic AI in solution design, and AI-driven contact centres and robotics in back-office processes.
Standard Bank is also applying AI within its software engineering practices, including the reading, writing, testing and deployment of code.
Fisher says these initiatives have translated into tangible gains, including improved operational efficiency, enhanced sales intelligence, more automated workflows and increasingly personalised products and services.
Absa’s Benvenuti comments that there are many complex tasks that AI can do so much better than humans, such as processing large amounts of information, noticing unusual patterns in a client’s transactional behaviour, and putting together various pieces of information from disparate sources to unlock insights into the client.
“We’ve had particular success in our financial crime operations, where AI has removed the burden of having to gather, collate and interpret data, allowing more data points to be used to assess risks posed by certain behaviours, which has allowed investigators to focus on more complex tasks, resulting in higher quality risk mitigation. Insights into client behaviours will also translate into better products and better service to our clients.”
Share