Financial services has spent years teaching people to use digital tools. The harder question now is whether they know when to trust them, when to challenge them, and how to turn them into better decisions.
Financial services has become very good at putting new technology into people’s hands. Dashboards, workflow automation, data platforms and generative AI are moving quickly from innovation teams into everyday work. Employees learn how to navigate a system, build a report, automate a task or write a better prompt.
That matters. But it is not the same thing as being digitally competent.
A person can be fluent in a tool and still make a poor decision with it. They can produce an impressive AI-generated client summary without noticing that the underlying data is incomplete. They can automate a process and remove a control that mattered. They can accept a model output because it looks precise, even when the context says it deserves a second look.
The question is no longer simply, “Can our people use the technology?” It is, “Can they use it with enough judgement to create value without creating new risk?”
Fluency gets you to the tool. Competence gets you to the decision.
Digital fluency means people are comfortable with technology, understand the basic language around it and can work with the tools available to them. In a bank, that might mean a relationship manager using an AI-enabled platform to identify patterns in a client’s transaction history, or an operations team using automation to reduce repetitive processing.
Digital competence starts one step later.
The relationship manager still has to decide which insight is relevant, which is noise, and how to turn it into a conversation that strengthens trust. The operations team still has to understand what happens when an exception falls outside the automated rule set. Someone has to ask whether the process is fair, compliant and actually better for the customer.
Those are not software skills. They are judgement skills exercised in a digital environment.
The World Economic Forum’s 2025 work on AI in financial services found that 90 percent of leaders believed their organisations needed significant adjustments, or a complete transformation, of their reskilling strategies to support the future. Its broader point is important: technology and data are not enough. Talent, culture and the ability of people to work effectively with AI are part of the implementation challenge.
Automation can remove effort. It cannot remove accountability.
Financial services is not an ordinary operating environment. Decisions can affect access to credit, fraud detection, insurance outcomes and customer data. Efficiency matters, but so do governance, fairness and the ability to explain a decision.
The Bank for International Settlements has highlighted both the opportunity and the risk. AI is being applied to areas such as data analysis, forecasting, payments and supervision, while institutions must also manage risks linked to data security, model behaviour, transparency and reputation. Its Project Noor work notes that AI models can influence decisions such as mortgage approvals, card limits and fraud flags.
That is why “the system said so” is not a satisfactory operating principle.
As more routine analysis is automated, human judgement becomes more valuable. People spend less time producing the first answer and more time interrogating it. Is the output plausible? What is missing? Does the recommendation make sense for this customer, in this context, under this regulation? When should a person override the automated path?
A digitally competent workforce knows that speed is not the same as quality, and confidence is not the same as correctness.
Digital leadership is now a team-level capability
There is another assumption worth challenging: that digital leadership belongs to the CIO, the transformation office or a small group of senior executives.
In practice, transformation is experienced much lower down the organisation.
It happens when a team leader decides how a new workflow will be used, when a business analyst translates strategy into a workable process, or when a compliance specialist questions a new use of customer data. It also happens when an operations manager spots an unintended customer outcome and escalates it.
That is digital leadership too.
The best digital organisations have people at multiple levels who can connect technology to operational performance, customer needs, regulation and sound decision-making. Adoption is not the finish line. Application is.
Training the hand without training the head
There is a cost to getting this wrong.
Tool-only training can create a workforce that moves faster without necessarily thinking better. Outputs multiply, reports arrive sooner and automated decisions scale. Yet weak judgement also scales.
In financial services, small errors can travel a long way. A questionable assumption used once is a mistake. Embedded in an automated workflow, it can become a repeatable operating problem.
The answer is not to slow digital adoption or make every employee a technologist. It is to broaden digital capability. People need technical fluency, but also problem-solving, ethical awareness, customer judgement, data literacy and the confidence to challenge an output when something does not look right.
This is the problem CPS was trying to solve when it developed its Digital Innovation: Transformation skills programme at NQF Level 6. The programme is built around practical, ethical and strategic application, including digital and data-enabled problem-solving, ethical leadership and customer-focused innovation, rather than treating digital development as tool training alone.
The next skills gap is not where many organisations are looking
Banks and corporates will keep investing in AI. They should. The productivity opportunity is significant, and customers will increasingly expect faster, smarter and more personalised experiences.
But the organisations that gain the most from these tools may not be the ones with the highest number of AI users. They may be the ones with the highest number of people who know what good use looks like.
Instead of asking how many employees have completed AI training, ask what they can now do differently. Can they solve a real operational problem? Spot a weak output? Explain a technology-enabled decision? Balance speed with control? Recognise when human intervention is still necessary?
Those questions are harder to measure than course completion, but much closer to the outcomes transformation programmes were meant to deliver.
On 8 October 2026, CPS’s Future Skills Forum will explore this shift through the theme, “How Banks and Corporates are Redefining Skills for the AI Era.” One useful place to begin is with a simple distinction: using digital tools is now expected. Knowing how to apply them with judgement is where the real capability begins.
Perhaps the skills gap is not that people cannot use the technology. Perhaps we have been training for the wrong finish line.
