There is a quiet assumption running through many conversations about AI, that as automation increases, the need for management will reduce. Fewer layers, flatter structures, more efficiency. On paper, it makes sense. In practice, something very different is happening. Managers are not disappearing. They are being compressed.
Every major technological shift redistributes work. Tasks move, roles evolve, responsibilities stretch. AI is no different. Execution is becoming faster, more automated, and in some cases less dependent on human input. But as that happens, something else expands at the same time. Oversight, judgement, decision-making, and accountability do not disappear into the system. They concentrate, and they concentrate most heavily in the managerial layer.
Strategy continues to flow down, execution flows up, and AI accelerates both. Managers sit in the middle, expected to translate, interpret, and hold everything together. Not as a redundant layer, but as the point where complexity lands.
Historically, managers acted as translators between strategy and delivery, supporting teams while maintaining direction and performance. That role has not gone away; it has expanded in multiple directions at once. Managers are now expected to oversee AI-augmented workflows they may not fully understand, maintain ethical and governance standards in rapidly evolving values and regulatory systems, support teams navigating constant change, and deliver results at increasing speed, often under already stretched conditions.
At the same time, span of control is widening, support structures are thinning, and expectations are rising. Gartner suggests that 75% of organisations expect managers to take on significantly broader responsibilities as AI adoption increases. And it is happening faster than most organisations are accounting for.
What makes this more significant than it first appears is that managerial strain is often misdiagnosed as a capacity issue. Too much work, not enough time. But what is happening here is more structural. Managers are becoming the shock absorbers of organisational change, holding the tension between competing demands. They absorb ambiguity from leadership, pressure from targets, uncertainty from teams, and increasingly, the complexity introduced by AI itself.
When that layer begins to fracture, the impact is not contained. Decision quality starts to slip, governance weakens, teams lose clarity, and trust begins to erode. Performance begins to drag. What looks like isolated burnout can quickly become something more systemic, something that, in many cases, could have been anticipated earlier.
There is another tension emerging alongside this. As AI takes on more execution, the relative importance of human judgement increases. Decisions about how AI is used, where boundaries sit, how outputs are interpreted, and what risks are acceptable. These are human questions, and they sit squarely with managers.
Yet many organisations are investing heavily in AI capability while underinvesting in the development of judgement, ethical reasoning, and decision-making under complexity. The result is a growing gap between what the system can do and what the people responsible for it are equipped to manage.
When energy is depleted, they feel it first. When alignment drifts, they are the ones trying to stabilise it. When complexity increases, they are expected to make it workable. They are not just another layer in the organisation. They are the leadership spine.
This is where visibility becomes critical, but not in the traditional sense. It is less about tracking workload or engagement in isolation, and more about understanding how pressure is experienced across managerial layers. Where span of control is stretching beyond what is sustainable, where accountability is increasing without corresponding support, and where perceptions of fairness and pressure are starting to shift.
At WellWise, we enable organisations to uncover surface these forms of strain earlier, before they show up as burnout, governance failure, or performance breakdown. In many cases, the signals are already present, they are just not being interpreted in time.
Alongside this, leadership development itself is being reshaped. Less emphasis on traditional management capability, and more focus on operating effectively in complex, AI-augmented environments. Strengthening judgement, clarifying decision rights, and ensuring that accountability is matched with support are becoming far more central.
Because as AI accelerates execution, the margin for error narrows. In that environment, the strength of the managerial layer becomes a defining factor in whether organisations scale effectively or begin to fracture under pressure.
The extinction curve is not about managers disappearing. It is about whether the role, as it is currently designed, can withstand what is coming next. And for many organisations, that question is still unanswered.