African Banks Deploy AI at Scale; Financial Sector Leads Continent's Tech Shift

African banks are already past the proof-of-concept stage. At the second CNBC Africa AI Summit, held Thursday at the Sandton Convention Centre, executives from the continent’s largest financial institutions, infrastructure operators and industrial groups arrived with deployment data, not forecasts.

The financial services sector produced the most striking numbers. Discovery Bank reported that more than half of all client interactions now flow through its in-app AI channel, a transition completed in just three to four months. Hylton Kallner, the bank’s chief executive, said the channel operates in natural language with utility equivalent to human banker interaction, and client satisfaction scores rank as the bank’s highest on record. Kallner framed the economic value as democratisation of premium banking services across the client base, not concentration at the top end of the market. He positioned South Africa’s market structure as a proving ground with global relevance, citing first-world financial sophistication, developing-market reach, regulatory credibility and high consumer technology adoption. “When you bring the two together, you get this amazing combination, this reaction that you see in Africa and virtually nowhere else in the world,” he said.

Additional reference context is available at https://www.cnbcafrica.com/2026/from-ambition-to-operating-reality-ai-summit-shows-african-business-already-building-the-future/.

Absa Group’s Kenny Fihla described AI already embedded in client onboarding, query resolution and software development, with most code now developed and launched by AI systems. The cost advantage flows to customers through cheaper service access, aggregation and reconfiguration, with Fihla projecting significant reduction in cost to serve.

Nedbank put a number on it. The bank quantified its AI rollout value at 375 million rand. Portia Matsena, divisional executive for hyper automation, said fraud case capture time fell from 30 to 40 minutes down to two minutes. Branch tellers freed from routine administration were redeployed into sales roles, lifting sales output from that cohort to 13 percent from 5 percent.

Meanwhile, the summit’s broader argument was about who captures the returns. The opening keynote set an ambition that threaded through subsequent sessions: Africa should shape the AI economy rather than simply consume it. Gauteng MEC for Economic Development Vuyiswa Ramokgopa told delegates the province has secured approximately 206 billion rand in investment commitments across 90 bankable projects at the 2026 Gauteng Investment Conference, part of more than 500 billion rand mobilised over two years. Her framing centred on positioning young people as builders, contributors and owners of the technology.

Dr Bongani Andy Mabaso, group chief technology officer at Altron, argued the continent does not need to build every layer of the technology stack to participate meaningfully. He advocated selecting strategic layers, starting with data. “The key to what makes us valuable is our own cultures, our traditions, which is contained in our data sets,” Mabaso said. He pointed to Africa’s projected rise to the world’s second-largest population by 2035 and to an export opportunity for solutions built cheaper, faster and matched to specific problems. Professor Letlhokwa Mpedi, vice-chancellor of the University of Johannesburg, added the continental scale argument: negotiating as 54 countries and 1.4 billion people changes the terms of engagement with global technology providers fundamentally.

Infrastructure operators presented the case that AI’s biggest African dividend may come from the physical economy. Sabine Dall’Omo, chief executive of Siemens sub-Saharan Africa, described AI as a new operating system for utilities and infrastructure, with municipalities already extracting value in predictive maintenance, call logging and scheduling. Her most concrete example was logistics: applying AI scheduling to coal flow through Richards Bay coal terminal alone, she said, would significantly grow South Africa’s GDP by optimising productivity and increasing product throughput. “Because we don’t have some of the traditional hardwired infrastructure, we can leapfrog into the next level, into the AI-enabled infrastructure,” Dall’Omo said.

Steven Santini, Schneider Electric’s secure power vice president for sub-Saharan Africa, noted modernisation investment is already visible in Johannesburg and Cape Town and spreading outward, with private microgrids proliferating in mining, industry and shipping ports. His counsel was disciplined sequencing, captured in a phrase that resonated with delegates: “A lot of it is slow down to speed up.”

Telkom’s Naeem Seedat reported 99.9 percent network uptime built on machine learning that predicts faults before they occur. His strategy focuses on what he called the AI last mile, connecting every person, business and entrepreneur to the AI economy at the lowest possible cost.

The workforce data offered the day’s most optimistic signal for operators weighing adoption risk. PwC Africa chief executive Dion Shango said research shows nearly two-thirds of workers are already using AI, with Africans in many cases ahead of their employers in adoption. Combined with the world’s youngest population, Shango argued, that enthusiasm represents a structural advantage. “If young people are that one cohort that are first to adopt the technology, then surely that puts Africa in a much more advantageous position than most other continents,” he said. PwC’s own experience backs the productivity case: a sustainability reporting tool built by one of the firm’s teams reduced work that previously took two weeks to a couple of hours.

Tertius Zitzke, group chief executive of 4Sight Holdings, described a client that automated 17 accounts payable roles end to end and redeployed the team into sales rather than retrenching them. Eighty percent of those people remained employed in new positions, skilled up and growing the business. At Absa, Johnson Idesoh, group chief officer for information and technology, said the fastest unlock came from putting tools directly in leaders’ hands. “We saw a massive change when we sat our executive and our board down and gave them hands-on experience on AI,” he said.

CNBC Africa vice-chairman Sid Wahi, the summit’s founder, closed by shifting the question from adoption to execution. “Artificial intelligence is no longer a conversation about some distant future,” Wahi said. “It’s already influencing how businesses operate today and how capital is allocated across banking, energy, enterprise.” His parting challenge was direct: “The opportunity is significant, but it will not be captured by enthusiasm alone. It’ll be captured by those willing to do the difficult work of turning ambition into capability.”

The CNBC Africa AI Summit returns in 2027, with the question no longer being whether African institutions will deploy AI at scale, but how quickly the capital flowing into that deployment translates into durable competitive advantage.