What the numbers look like in African banking
When Backbase teamed up with African Banker to survey 277 banking executives across 37 countries for The State of AI in African Banking 2026 report, we wanted to see how this played out locally.
The responses revealed a major disconnect inside bank leadership teams:
- Finance teams track AI return on investment at 82%.
- C-suite executives track it at just 50%.
- Risk and compliance teams track it at 48.1%.
Finance tracks the numbers because tracking money is their core job. But in many institutions, AI funding sits in a separate innovation bucket instead of being reviewed like standard technology spending. When the C-suite isn't tracking those numbers, the board never sees them.
Why bigger banks struggle more
You might assume that larger, multi-country banks have better visibility into their technology spending. The survey showed the opposite.
Regional banks report tracking AI ROI at 90.9%, the highest of any group. Pan-African banks, which are usually the most advanced institutions on the continent, sit at 54.2%.
When a bank operates across multiple regulators, currencies, and legacy core systems, getting a single clear view of technology spending becomes much harder. Growth creates operational complexity, and that complexity hides whether a technology project is actually making or saving money.
Three practical questions for the next board meeting
Fixing this gap doesn't require a deep technical understanding of machine learning models. It requires basic financial discipline.
Board directors can reset the conversation by asking three simple questions:
- Where is AI currently changing a specific line item on our balance sheet?
- Which active AI projects have a clear financial metric, a starting baseline, and a named manager responsible for both?
- What projects lack those three things, and when will we either measure them properly or stop funding them?
As we discussed in the first article of this series on the African banking ROI gap, asking for clear returns isn't about slowing down innovation. It is about making sure the bank funds projects that actually deliver value.
Setting up the right technical foundation
A board cannot track numbers if the bank's underlying software hides the data. If a system cannot show why an automated decision was made or what business result it produced, reporting will always rely on guesswork.
This is a major reason why fewer than 10% of banks successfully scale AI into production. They try to add smart tools on top of messy, disconnected platforms. Whether a bank is automating simple customer requests or rolling out autonomous banking workflows, governance needs to be built directly into the underlying platform. The system should track financial impact and check compliance rules automatically, before an action takes place.
Common questions
1. What percentage of C-suite executives in African banking track AI ROI?
Only 50% of C-suite executives track AI ROI according to The State of AI in African Banking 2026 report, compared to 82% of Finance teams at the same banks.
2. Is the AI board oversight gap unique to African banking?
No. Global studies show only 15% of enterprise boards get regular AI performance reports, and 39% of Fortune 100 companies report having board oversight for AI. The issue exists worldwide.
3. How can bank boards ensure accountability for AI spending?
Boards can tie every AI budget approval to existing financial metrics, assign a single manager to be accountable for each project, and ensure their underlying technology platforms automatically track operational outputs.
