Financial leaders are now being asked to fund artificial intelligence (AI), govern it, and explain the value it creates. That is not a simple technology question. In banking, it is becoming a test of leadership.
AI is starting to influence decisions at the core of the business, for example, in credit, fraud, pricing, collections, customer engagement, compliance, and operational resilience. These decisions affect customers, regulators, shareholders, employees, and the institution’s ability to compete.
Speed alone will not define the next phase of AI in financial services. Banks will need discipline, especially as AI begins to influence decisions that customers and regulators may later challenge.
The investment is already happening. The SAS Data and AI Impact Report, with research insights from IDC, found that banks are ahead of other sectors in AI spending and in the adoption of trustworthy AI practices. That sounds encouraging, and it is, up to a point. The same research also shows a trust gap that still exists. Only 11% of banks have both high internal confidence in AI and systems that are demonstrably trustworthy. Nearly half fall into what IDC describes as the “trust dilemma”, either underusing reliable AI because they do not trust it enough or over-relying on AI that has not been properly validated.
Governance considerations
Confidence can be misleading. A model can perform well in testing and still fail the institution if the data is fragmented, governance is weak, or no one can explain how the decision-making process took place. Banking is not forgiving terrain for unclear decisions. The problem usually shows up in ordinary places, such as a credit recommendation that cannot be properly traced, a fraud model that produces too many false positives, or a customer receiving inconsistent treatment across channels. The issue is not only whether the model works but whether the bank can explain, monitor, and adjust the decisions it supports.
Governance cannot serve as the final stamp of approval at the end of the innovation process. It has to be part of how AI is designed, deployed, and managed from the start. This does not mean slowing everything down. In many cases, good governance helps organisations move faster because teams know the rules. They know which data can be used, which models need review, which decisions require human oversight, and where escalation is needed.
In a regulated environment, that clarity is what allows innovation to survive real customers, regulators, and market pressure.
Getting value
Return on investment (ROI) needs the same discipline. Efficiency matters, and banks cannot ignore cost pressure. But the stronger AI business case is not always found in replacing effort. It is often found in improving the quality, speed, and consistency of decisions.
The SAS/IDC banking findings support this. Organisations using AI to improve customer experience reported stronger returns than those focused primarily on cost savings. The study also found that organisations prioritising trustworthy AI were 60% more likely to report doubling overall return on their AI initiatives.
Instead of asking only what can be automated, banks should ask where better decisions can create measurable value, faster and fairer credit assessment, and more accurate fraud detection, among other benefits.
For African financial institutions, the stakes are immediate. Banks are operating in markets shaped by digital acceleration, regulatory scrutiny, fraud pressure, affordability constraints, legacy systems, siloed data, and rising customer expectations. AI has a role to play, but only if the environment around it is strong enough to support it.
An integrated approach
This is where financial leaders play a critical role. AI cannot belong only to data science teams or technology functions. It has to connect with finance, risk, compliance, operations, and customer strategy.
The banks that make progress will bring these functions closer together around shared decision-making. Agentic AI makes this even more important. AI agents can analyse data, make decisions, and take action across workflows with limited human intervention.
In an enterprise setting, AI agents need more than language models. They require trusted data, advanced analytics, decision logic, governance, and compliance to deliver reliable, auditable outcomes.
Banks will use more advanced AI. What is less clear is whether leadership has created firm enough boundaries around where it may act, when it must escalate, and who remains accountable.
Delivering value
For African banks, the opportunity is to build AI into the institution with discipline from the beginning: cleaner data, clearer rules, measurable outcomes, and people who remain accountable.
AI will help banks move faster. The harder task is making sure speed does not weaken accountability. In financial services, leadership still comes down to the quality of the decisions made, the evidence behind them, and the willingness to stand behind them when it matters.
