By Dilys Sillah

Every conversation about artificial intelligence eventually arrives at the same promise: it will save time. Tasks that once took hours can be completed in minutes, routine work can be automated and employees can be released to concentrate on activity of greater value. It is an attractive proposition, particularly for organisations under pressure to reduce costs, improve productivity and do more with fewer resources. Yet beneath that promise sits a question that is receiving far less attention: once AI has saved the time, what determines whether people can turn that time into meaningful value?

The assumption appears to be that the answer is self-evident. If an employee has eight hours returned to them, the organisation has gained eight hours of productive capacity. If AI removes a repetitive task, the person who previously performed it can simply move on to more complex, creative or strategic work. On paper, the arithmetic is compelling. Human beings, however, do not operate like entries on a spreadsheet, and time made available is not the same as human capacity restored.

This is where an important distinction needs to be made between released capacity and human capacity. Released capacity is the operational space that AI appears to create. It may be expressed as hours saved, tasks removed, faster output or a reduced requirement for human labour. It is visible, countable and relatively easy to place inside a business case. Human capacity is something different. It is the changing condition that determines whether a person can deploy the knowledge, skills and experience they already possess effectively, consistently and sustainably.

A person may be highly capable and still have limited capacity. They may understand the technology, have completed the training and be entirely willing to adapt, while also being cognitively overloaded, emotionally depleted, unclear about what is expected of them or working within a culture that makes experimentation feel unsafe. Their capability has not disappeared, but their ability to use it well has been affected. Giving that person an AI tool, or returning several hours to their week, does not automatically address the conditions that are constraining their performance.

My work through The Glass Framework™ begins from the understanding that capacity cannot be separated from the condition of the human being. It considers depletion, presence and wholeness because what people are able to bring to their work is shaped by far more than their technical competence.

Human capacity is not a fixed container that is either full or empty, nor is it simply another word for skill. It can expand, contract and become obstructed by the conditions in which a person is expected to perform. This matters in any programme of change, but it becomes particularly important when organisations introduce technology that alters not only how work is completed, but also what remains for people to do.

AI does not always remove pressure. In many cases, it changes the nature of the pressure. An employee may spend less time producing a first draft, yet more time checking whether the information is accurate. A manager may receive faster analysis, yet face a greater volume of decisions. A team may automate routine processes, only to find that the work left behind requires more judgement, concentration and accountability. Roles can quietly expand before job descriptions, authority and support arrangements catch up. What looks like released capacity from the organisation’s perspective can therefore be experienced as increased complexity by the person doing the work.

The emerging evidence reflects this tension. Boston Consulting Group reported in 2026 that 42 per cent of regular frontline AI users were saving at least one working day each week, yet 66 per cent received limited or no guidance about what to do with the time saved. The same research found that 41 per cent experienced increased cognitive load.

BearingPoint subsequently reported that almost three-quarters of organisations implementing AI could identify measurable financial value, while only 13 per cent had scaled their initiatives in line with the original business case. It also found that many organisations were already reporting workforce overcapacity created by AI, while still struggling to decide how that capacity should be redeployed.

These findings do not suggest that AI is failing. They suggest that organisations are measuring one part of the transformation and assuming the rest will follow. Time saved is being treated as an outcome when it is only an input. The real outcome depends on what happens next: whether roles are redesigned intelligently, whether leaders provide direction, whether employees understand how their contribution is changing and whether the organisation has created the conditions in which people can absorb greater complexity without becoming depleted by it.

The distinction also changes how we should think about AI readiness. Readiness is often discussed in terms of infrastructure, data, governance, technical skills and access to tools. All of these are essential, but they do not tell us whether the people inside an organisation are in a condition to adopt, adapt and use AI effectively.

Training can increase capability. It cannot, by itself, resolve exhaustion, mistrust, role ambiguity, poor management or the cumulative effects of constant organisational change. If those conditions remain invisible, leaders may interpret inconsistent use as resistance or a skills deficit when the deeper problem is depleted human capacity.

This is especially relevant across Africa, where the urgency to participate in the AI economy is understandable and necessary. Governments and businesses are being encouraged to invest in infrastructure, digital skills and locally relevant technology. Yet effective adoption will not be secured by importing tools and training people to operate them. Organisations must also understand the cultural, institutional and human conditions into which those tools are being introduced. The question is not simply whether people can use AI. It is whether they can use it with judgement, confidence and sufficient presence to produce outcomes that are valuable in their own context.

For leaders, this requires a different set of questions. It is no longer enough to ask how many employees have received licences, how many have completed training or how many hours a system claims to have saved.

Leaders must ask what new work has replaced the work that disappeared, whether decision-making and verification demands have increased, whether people know how to use the time released and whether managers are equipped to lead teams whose roles are changing faster than organisational structures.

They must also resist the temptation to interpret every difficulty as reluctance to change. A capable employee who is not using AI consistently may not be unwilling or unable. They may be signalling that the surrounding conditions are not yet sufficient for effective utilisation. That is why human capacity must be diagnosed rather than assumed. Without that understanding, organisations risk investing repeatedly in tools, training and process redesign while leaving the condition that determines effective deployment untouched.

The organisations that gain the greatest value from AI will not necessarily be those that introduce the most technology or calculate the largest number of hours saved. They will be those that understand the difference between making capacity available and enabling human beings to use their capability well. AI can release time, remove tasks and widen what is technically possible. It cannot, on its own, restore a depleted workforce, create clarity, build trust or make people whole.

That work remains a leadership responsibility. Until organisations recognise it, the gap between AI deployment and effective utilisation will persist, and the productivity they expected to gain may continue to exist largely on paper.

AI can release operational capacity. Whether that becomes sustainable value will depend on the human capacity available to receive it.

Sillah is a human development practitioner, communication specialist and executive leadership coach. She is the creator of The Glass Framework™.