While AI can make individual tasks faster, it does not automatically improve entire processes. It is time for businesses to invest in employees who can bridges gaps between AI and day-to-day operations, says Paul Soares
IMAGINE a firm that produces a client-review summary in seconds, yet takes just as long to complete the review. Missing records still have to be chased. Conflicting information still needs resolving. Nobody has clarified who owns the next approval.
Drafting is faster. The client is still waiting.
This hypothetical example exposes a leadership problem. The technology has improved a task, but nobody has changed the way the work moves through the business. Someone needs to connect those two things. The supplier understands the technology; the business understands its clients, controls, exceptions and operational reality. The scarce capability is the person who can translate between the two.
For Jersey’s financial-services firms, that person may already be working in operations, client service, risk or another part of the organisation. The question is whether leaders recognise the potential and create the conditions for it to develop.
Jersey’s regulator has made a similar point. The JFSC’s AI guidance, published in July, created no new rules, but it was clear that firms remain accountable for how they use AI and need “sufficient people with the required skills and capabilities” to oversee it.
The Jersey AI Playbook names the skills gap as a critical barrier, and not only for technical staff: directors and leaders need AI literacy too.
Look for judgment, not just enthusiasm
I would start by looking for people who understand a process, can explain why its controls exist and are willing to question how it operates. Who can identify the exception missing from the procedure? Who can distinguish a necessary safeguard from a workaround that has outlived its purpose?
An experienced investment-management, trust or fund-administration professional may know why particular information needs checking, which client circumstances change an approval and where responsibility becomes unclear. That knowledge is valuable because a technology project needs an accurate account of the business, not just its process map.
But familiarity is only part of the requirement. The person also needs curiosity, judgment and enough confidence to challenge both the existing process and the proposed solution.
The aim is not to nominate the longest-serving employee or the most enthusiastic user of an AI tool. It is to identify people who can translate operational reality into questions that specialists can actually solve.
Augment experience – do not simply relabel it
This is why, after more than two decades in financial services, I completed the London School of Economics’ AI Leadership Accelerator. I was not trying to replace that experience. I wanted to extend its usefulness: to understand what could be commissioned, challenge assumptions and recognise where specialist help was needed.
I studied alongside a full-time role and that taught me something about the rest of this argument: development squeezed into evenings and weekends is hard to sustain. Firms should not rely on it.
For experienced professionals, adapting does not have to mean abandoning what made them valuable. It can mean adding enough new knowledge and practical discipline to make their judgment more useful in a changed environment.
But development is not relabelling. Years in financial services do not automatically confer AI competence, just as formal study does not prove an ability to deliver working change. Technical specialists remain essential. The opportunity is to create people and teams capable of working across those boundaries intelligently.
Change the job, not only the job description
Identifying a capable employee is the beginning of the leadership decision, not its completion. What will be removed from that person’s existing workload? Which responsibilities will be reassigned? Who will give them access to the colleagues, information and specialist support the work requires?
Protected time needs to mean time that survives contact with everyday priorities. A new responsibility without time or authority is a workload increase, not a capability strategy.
The mandate also needs to cross the boundaries the work crosses. Someone asked to improve an end-to-end client process cannot do so if they may examine only their own department’s contribution. An executive sponsor should help resolve competing priorities and make clear who can decide what changes.
This is not a cost-free initiative because the people are already employed. Leaders are allocating capacity to a different purpose.
They should make that choice explicitly rather than expecting development to happen around an unchanged job.
Give learning somewhere to work
Start with one material but bounded process. Give the team an outcome to improve – a review completed with less rework, for example – rather than a general instruction to “find uses for AI”. Ask them to observe how the work actually happens, including the exceptions and informal checks. Then bring operational, risk and technical colleagues together to examine what should change.
Learning becomes useful when it helps the team frame a better problem, challenge a proposed answer and explain why a particular approach should or should not be adopted.
Development should also be relevant to the role. The person redesigning a workflow needs different depth from the employee using it or the manager supervising its operation. Attendance at the same course does not establish that all three can do what the business expects.
Managers are part of this investment. They need to support challenge and distinguish faster work from better work. If an employee is told to exercise judgment but rewarded only for clearing cases quickly, leadership has created conflicting expectations.
The test of development should therefore be application: can the team explain what changed, why it changed, what still requires specialist input and whether the intended business outcome improved?
Make capability a continuing responsibility
The work does not end with a course or a first project. Technology changes; so do business processes, regulation and the organisation’s own risk appetite. Someone needs an explicit responsibility for keeping the connection between those worlds current. Nor should the capability remain an informal dependency on one enthusiastic individual. A clear remit, shared learning, documented decisions and continuing access to specialists turn individual development into organisational capability.
For Jersey, the opportunity is constructive but conditional. Accumulated experience need not become obsolete; neither can its future usefulness be assumed. It becomes more valuable when people are given the opportunity to extend it and the means to apply it.
The people may already be here. Recognising their potential is only the beginning. The leadership question is:
Who should we develop, and what must we change to enable them?
