Spend first, measure later: Inside African banking's AI boom

The African banking sector is buying into AI like a new drug, but the dosage is still a mystery. 277 executives across 37 nations have pledged to lift their AI spend by the next year, yet almost a third of them are not even measuring the return. 82 % of those blind investors will still expand budgets. The result? A continent racing toward automation while its balance sheets stay on faith.

Standard Bank’s commercial asset finance division cut loan‑vetting time from an hour to a minute. That 60‑second turnaround, highlighted in its March 2026 Capital Markets Day filing, is a textbook case of AI delivering speed. The bank’s conversational bot now handles 65 % of digital queries, and by August 2026, 39 000 employeesabout 72 % of its workforcewere regular users of generative AI. For staff, the promise is clear: administrative backlogs vanish, customer friction drops, and human talent can focus on high‑touch advisory roles. For the bank, however, the financial dividend remains hazy. Every line of code, every data‑governance protocol, every integration layer carries a price tag that is hard to reconcile with short‑term profits.

Deploying AI in regulated banking is not a plug‑and‑play affair. It demands rigorous data governance, expensive software licences, and complex systems integration. Legacy core banking systems act as bottlenecks, forcing banks to spend on cloud hosting and on‑premises infrastructure. The shortage of machine‑learning engineers across the region only adds to the cost curve. In this environment, the promise of automation is weighed against the weight of compliance and integration overheads.

A joint 2026 report by Backbase and African Banker surveyed 277 executives. While 83.2 % plan to increase AI spending, only 67.1 % of institutions actually track ROI. Among those that do not track, 82 % still intend to expand budgets. The disconnect is stark: banks are betting on AI as a strategic priority, yet the mechanisms to prove its profitability are largely absent. The data shows an uncomfortable divergence between capital expenditure and performance tracking that could spell trouble if left unchecked.

Everyone’s celebrating AI as the next big growth engine for African banks, but the reality is that most institutions are still operating in a “spend first, measure later” mode. This approach mirrors the early days of mobile banking in the 2000s, where banks invested heavily in technology without a clear revenue model, only to later discover that the real value lay in customer acquisition and retention, not in the tech stack itself. If banks repeat that mistake, they risk inflating costs without delivering proportional returns.

By 2028, I predict that at least 40 % of African banks will abandon or scale back AI projects that fail to deliver measurable ROI. Those that survive will have built robust measurement frameworks, integrating AI metrics into their core financial reporting. The shift will be driven by the need to satisfy investors, regulators, and a growing cohort of tech‑savvy customers who demand transparency.

The AI boom in African banking is still a gamble. Speed and operational efficiency are real, but without a clear path to profitability, the continent’s banks risk turning innovation into a costly vanity project.