Why Adding AI Can Lead to More Work
Many companies have adopted generative AI only to find that workloads have barely changed. In some cases, employees are now spending additional time managing new AI tools or reviewing AI-generated outputs.
Generative AI should be a powerful means of driving business transformation by redesigning how organizations operate and how work gets done, while freeing people to spend more time on higher-value activities. But when AI adoption itself becomes the objective, the impact often remains limited to incremental efficiency gains: preparing materials faster, drafting emails more easily, or summarizing documents with less effort.
A more effective approach is to structure generative AI adoption in two stages: first determine what can be eliminated, and then decide where AI can expand the organization’s capabilities. In practice, however, many companies reverse this sequence. They begin by asking what AI will enable them to add rather than what existing work can be removed. The result is often an accumulation of new activities on top of existing ones. New tools are introduced, creating work to manage them. More data can be analysed, generating more reports to summarize the findings. More information becomes available, leading to more meetings to share it.
That is why successful AI-driven transformation should not begin with the question, “What more can we do with AI?” It should begin with, “What should we stop doing?”
Deciding What to Stop Is a Leadership Responsibility
The process should therefore begin by taking stock of existing work. Meetings, forms, reporting processes, approval flows, and other activities should be reviewed systematically for their necessity, the value they create, and the resources they consume. Most meetings, for example, serve one of three broad purposes: sharing information, discussing issues, or making decisions. Information sharing can often be handled through email or documents. If discussions repeatedly produce no meaningful outcome, the purpose of the meeting itself may need to be reconsidered. And if a meeting consistently fails to reach a decision, it may not be structured to support effective decision-making. The review should ultimately determine whether each meeting serves a clear purpose and creates sufficient value to justify the time and people involved. Where it does not, eliminating the meeting should be an option.
Crucially, however, decisions about what to stop should not be left entirely to the teams performing the work. From within each function or project, there are often legitimate reasons for continuing existing activities. Sales has its priorities, administrative functions have theirs, and both Project A and Project B may have valid reasons to argue for more resources. Asking each team simply to retain what it considers necessary can therefore result in almost everything being justified and very little actually being eliminated. What makes sense within an individual function or project does not necessarily make sense for the organization as a whole. Deciding what to retain and what to stop requires strategic decisions about the organization’s overall direction and the allocation of resources across it. That responsibility ultimately sits with leadership.
Companies may attempt to eliminate unnecessary work by reorganizing teams or functions, but changing the organizational structure alone will not change how work gets done. If the underlying day-to-day activities remain unchanged, the result may simply be a new organizational chart and new reporting lines, while unnecessary reports, meetings, and other routines continue as before. Meaningful business transformation therefore requires companies to redesign the work itself, eliminating activities that no longer justify the resources they consume and creating capacity that can be redirected toward higher-value activities.
Trying to Satisfy Everyone Can Mean Deciding Nothing
Deciding what to stop, however, is rarely easy. Eliminating an activity may create resistance from those who have been responsible for it. Prioritizing sales may put pressure on margins, while protecting margins may disappoint customers in the near term. When both sides have legitimate arguments, leaders can be drawn toward solutions that try to preserve both priorities or secure broad agreement. But attempting to accommodate every interest can ultimately delay the strategic decisions the organization needs to make.
Consider a situation in which an organization has two projects, A and B, but needs to prioritize A. Doing so may require Project B to be scaled back or stopped altogether, potentially resulting in losses. Yet if A is more important to the organization’s future, resources need to be reallocated accordingly. Giving equal priority to both projects would spread people, budgets, and time across them, leaving neither with sufficient resources to deliver its full potential. What appears in the short term to be the less disruptive choice may ultimately weaken the organization’s long-term competitiveness.
The point is not that leaders should be less considerate. The problem arises when consideration for immediate stakeholder concerns takes precedence over the decisions leadership is responsible for making. While teams are necessarily focused on immediate priorities and near-term delivery, leaders also need to look six months, a year, or further ahead. Their role is not to eliminate competing interests, but to take a longer-term perspective in deciding which of the competing options to prioritize and which to forgo, and to take responsibility for the consequences. Consideration has its place not in avoiding difficult decisions, but in how those decisions are communicated and implemented.
Take workforce reductions as an example. If retaining the workforce is the choice that best supports the long-term strength of the business, then that is a rational decision. But if a necessary decision is postponed primarily out of fear of short-term resistance, that should not be confused with consideration for employees. A business decision should not be judged primarily by whether it appears considerate or tough. What matters is whether the decision strengthens long-term customer value and supports the sustainability of the business and organization. Allowing short-term resistance or the desire to avoid unpopular decisions to override long-term business considerations can ultimately undermine value.
Turning AI Efficiency Gains into Higher-Value Work
Only after deciding what to eliminate can companies turn to where AI can expand the organization’s capabilities. This means deciding where to redirect the capacity that has been freed up, whether in the form of time, talent, or budget, to build a stronger business. Suppose generative AI reduces a task that once required 100 hours a month to 95. Those five hours may contribute to cost savings, but they also create capacity that can be redirected toward deepening customer understanding, testing new services, advancing R&D, or higher-value decision-making. It is through such strategic resource allocation that operational efficiency gains from AI can ultimately translate into business value.
Starting with how to use AI can easily pull companies toward what the technology makes possible. The more important task is to determine what to eliminate, where to concentrate resources, and what will make the business stronger. Only then should companies decide where AI can best support those priorities. The sequence matters.
Generative AI will continue to expand what companies are capable of doing. They will be able to analyse more information, test more initiatives, and create more products, services, and outputs. But as the range of possibilities expands, trying to pursue them all risks spreading resources too thinly. Effective leadership in the AI era is therefore not about knowing more than anyone else about what AI can do. It is about being clear not only about what to pursue, but also what not to pursue. Whether that freed-up capacity ultimately translates into future competitiveness depends on where companies choose to redirect it. That choice belongs not to AI, but to leadership.
