Johnson Idesoh, Absa’s group chief information and technology officer, says that artificial intelligence—whether generative, agentic or weaponised by cyber‑criminals—dominates his agenda, but he repeatedly stresses that AI must be anchored to a clear business strategy and measurable customer outcomes to be worth the investment.
In a candid conversation on TechCentral’s Meet the CIO podcast, the seasoned technologist warned that the rapid maturing of AI in heavily regulated sectors such as banking is a double‑edged sword: the same tools that can unlock new revenue streams are also being co‑opted by hostile actors. “If we do not use AI to hunt the threats that are already employing it, we will fall behind,” Idesoh told the host, underscoring the urgency for South African banks to turn AI into a defensive asset as well as a growth engine.
Idesoh’s résumé reads like a roll call of African‑wide financial and pharma powerhouses. Before joining Absa as group CTO, he steered technology strategy at Old Mutual, held senior roles at insurer Aviva and pharmaceutical giant AstraZeneca, and built a reputation for aligning complex IT landscapes with shifting corporate visions. His current mandate is shaped by three converging forces: a pan‑African, client‑led strategy championed by group CEO Kenny Fihla, an evolving regulatory framework, and the accelerating pace of AI, data analytics and cybersecurity innovation.
The banker’s most striking revelation concerns Absa’s R2.4‑billion software impairment forecast for the 2025 financial year—a thirteen‑fold jump from the previous year. Rather than a single massive write‑off, the figure represents the cumulative effect of over a hundred smaller asset devaluations. Idesoh explained that this reflects a deliberate shift away from legacy, siloed systems towards a more agile, cloud‑centric architecture that can better serve a continent‑wide client base.
How Absa’s AI strategy aligns with the new group direction
| Strategic Pillar | AI Initiative | Customer Impact | Risk Mitigation |
|---|---|---|---|
| Pan‑African client‑lead model | Agentic AI chat‑bot for account enquiries | Faster, 24/7 support for 30 million users | Reduces call‑centre load, lowers human error |
| Regulatory compliance | AI‑driven transaction monitoring | Real‑time fraud detection, AML alerts | Meets South African Reserve Bank (SARB) standards |
| Cybersecurity acceleration | AI‑enabled vulnerability scanning | Immediate patch recommendations | Counteracts AI‑powered attacks from cyber‑criminals |
| Talent development | Gamified AI skill‑upskilling platform | Upskilled staff, higher innovation rate | Builds internal expertise, reduces outsourcing |
The table illustrates how each AI thrust is deliberately tied to a business outcome, whether it be improved service speed, stricter compliance or hardened security. The takeaway is clear: Absa is not experimenting with AI in isolation; every deployment is a piece of a larger, customer‑centric puzzle.
Idesoh also reflected on the physical evolution of banking technology. From the water‑cooled mainframes of the 1990s, the industry has moved to compact IBM z16 machines and increasingly to cloud‑native services. Yet the real bottleneck remains the decades‑old COBOL code that still powers many core banking functions. “The hardware is cutting‑edge,” he remarked, “but the software paradigm is stuck in the past, and that’s where we need to focus our AI‑assisted refactoring efforts.”
When the conversation turned to the broader AI ecosystem, Idesoh addressed the buzz surrounding Anthropic’s “Mythos” model. He cautioned against panic‑driven reactions to AI‑generated vulnerability discovery, arguing instead for a steady, skill‑based adoption that enables security teams to scan, prioritise and remediate flaws at scale. In his view, the traditional junior‑to‑senior developer hierarchy may give way to apprenticeship‑style pathways, where AI tools accelerate learning and reduce the time to competency.
Absa’s internal AI champion, an IT‑support agent now handling queries from 11 000 colleagues with a 90 % resolution rate, exemplifies the practical benefits of agentic AI. On the customer‑facing side, a newly launched AI assistant can process routine banking requests, freeing human advisers to tackle complex, relationship‑driven work. Both agents are part of a broader talent‑development programme that rewards early adopters, celebrates new certifications through gamified leaderboards, and embeds a culture of continuous learning across the organisation.
Beyond the boardroom, Idesoh shared a personal anecdote that hints at his fascination with technology’s trajectory. As a teenager he dreamed of piloting aircraft, only to discover that the new generation of planes were essentially “flying computers”. That epiphany led him to his first programming experience on a BBC Micro, setting the stage for a career that now shapes AI policy for one of Africa’s largest banks.
The Meet the CIO episode finishes on a forward‑looking note: AI will not replace humans, but it will reshape the roles we play. Whether through apprenticeship models, AI‑augmented security operations or customer‑centric bots, the technology is poised to become a permanent fixture in South Africa’s financial landscape. As banks grapple with the twin pressures of regulation and digital disruption, the ability to translate AI hype into tangible value may well define the next decade of African banking.