The modern workday now begins before most people have even arrived at the office or opened their laptop, wherever they may be. Before the first conversation is had, the inbox is already swelling, the calendar has shifted and messages are gathering with that familiar insistence that everything matters at once.
AI was supposed to reduce friction in precisely this sort of environment, and, in some ways, it has, at least at the level of execution. But it has also made a deeper problem harder to ignore. Gallup’s 2025 Workplace Report describes a workforce that is straining beneath this acceleration: Only 21 percent of employees are engaged globally, 40 percent report experiencing stress “a lot” of the previous day and manager engagement has fallen to 27 percent.
Microsoft’s recent reporting on the “infinite workday” gives that strain a daily rhythm: Workers are interrupted roughly every two minutes, receive 117 emails and 153 Teams messages a day, and nearly half of employees, along with more than half of leaders, now describe work as chaotic and fragmented.
While it’s tempting to merely call this a problem of volume, the deeper issue is that work has become saturated with coordination while the human conditions that make coordination intelligent and effective have been left underestimated and poorly designed.
Communication channels multiply, but few are structured well enough to help people think and prioritize. Transparency gets confused with universal exposure, so everyone is expected to read everything, even when that only produces noise, delay and quiet mistrust. A thousand visible messages do not create clarity. More often, they create a low-grade panic in which nobody is quite sure what matters most, but everyone is expected to remain responsive.
Real trust is trusting others to sort out what is more important and let certain tasks be handled by certain individuals. As technology and communication channels develop, we need to choose a definitive path. Too often organiations act like a McDonald’s open around the clock, instead of a fine dining restaurant with better quality products but more limited opening hours.
The work of leaders today is to shape communication channels so that relevance governs attention, to create spaces in which mandates are clear, and to ensure that overlapping teams can coordinate without sliding into rivalry. Many managers agree to a great many new developments but tremble when it comes to properly implementing a direction. Here, it is imperative to incorporate the staff’s input and concerns while providing a space for them to constructively collaborate and contribute together as a team. Leaders need to involve them in the decisions rather than retreat from leading.
In the AI era, as more routine tasks are automated, the value of creative judgment, coordination and collaborative problem-solving will only rise. The task ahead for many organizations is to maintain the practical know-how required to review, repair and refine what automation produces. Rather than just reducing clerical burden, an increased reliance on AI may erode some of the quieter knowledge that used to be built through repetition.
What I see in organizations is not a shortage of intelligence, nor always a shortage of effort, but a shortage of environments in which intelligent people can work together without becoming defensive, territorial or exhausted. That is why I increasingly think of leadership less as the art of making faster decisions from above and more as the discipline of designing the conditions in which good decisions can emerge below and across and start to form before they ever reach the leader’s desk. For this to happen, the leader needs to define mandates clearly enough that people or groups know whether they are making the decision itself or preparing the basis for one.
That shift matters because the underlying problem is not only operational but motivational. When work changes quickly and people feel they have little influence over how it is being reshaped, engagement collapses for reasons that are more psychological than technical. The old triad of intrinsic motivation still matters: autonomy, competence and relatedness.
People need to feel that they have some meaningful say in how work is done, that they are still good at something that matters and that their work remains connected to others in an intelligible way. In the AI era, a badly handled rollout can erode all three at once. People lose not only confidence but also their sense of stake and direction.
In practical terms, that means four things.
- Redesign communication channels around relevance, not exposure. Not everyone should read everything, and calling that “transparency” often disguises a lack of trust in the organization’s own filtering. People need systems that help them distinguish what requires attention from what merely creates motion.
- Define mandates before you create the working group. One of the quickest ways to drain energy from people is to let them believe they hold meaningful responsibility, only for them to discover later that they were never empowered to decide anything at all. Leaders need to be explicit about whether a team is deciding, recommending or preparing the basis for a decision.
- Structure coordination so meetings produce innovation and a sense of joint responsibility, not self-protection. In a poorly designed environment, cross-functional groups begin to defend territory rather than solve problems. The task is not simply to get people in a room, but to create enough clarity and trust that overlapping expertise becomes useful rather than politically threatening.
- Involve people closely enough in change so that motivation rises rather than thins out. When people can shape how work is being done, rather than merely absorb what has already been decided, they tend to work with far more energy and seriousness. They feel closer to the mechanism of decision-making, and that almost always improves both commitment and judgment. It also relieves the leader from the position of being all-knowing in a fast-changing environment.
The leadership advantage in the AI era will not belong simply to those who automate more tasks or move more quickly through them. It will belong to those who can organize human judgment more intelligently—by reducing communication clutter, building trust in how information is filtered, defining mandates clearly and involving people closely enough in change that motivation rises rather than thins out. AI can accelerate work, certainly, but it cannot by itself produce coherence, trust or commitment. Those remain human achievements, and leadership, however much it mattered before, matters more now.
