AI will expose the managers who only report, not decide

You are in a leadership review where the manager appears fully prepared. The slides are clean, the numbers are current, the risks are colour-coded and every dependency has an owner. The update moves with confidence because nothing looks neglected. For a few minutes, the room sees what it expects to see: a manager in control.

Then the CEO asks a different question. What should we do now?

The manager pauses. He explains that there are three options. He refers to the dashboard, mentioning customer pressure, capacity constraints, and a few open dependencies. He says the team is evaluating the best way forward. The answer is professional, but the room has changed. The manager came prepared to report the situation. He was not prepared to own the judgment.

That is the managerial test AI will make harder to escape. For years, many managers created value by gathering updates, reconciling information, coordinating across teams, and presenting the current state with reasonable clarity. That work mattered when information was slow, fragmented and difficult to assemble. Senior leaders needed people who could bring the picture together.

AI weakens that protection. It can summarise status, detect exceptions, compare scenarios, generate first-order analysis, and expose patterns faster than most organisations can complete a review cycle. Once that happens, the manager who only brings information to the table faces a sharper question: if the system can describe the situation, what exactly are you here to decide?

The room now examines the case again. The customer wants faster commitment. Sales says the opportunity is important. Operations says the current plan may break quality. Supply chain says a supplier risk is rising. Finance says margin will suffer if the response is rushed. HR says the required capability is uneven across teams. The facts are not missing. The decision is.

This is where many managers get exposed. They do not fail because they are lazy or careless. They fail because the organisation has trained them to be custodians of status rather than owners of trade-offs. They know how to escalate, align, update, circulate, and wait. They are less comfortable saying which risk matters most, which commitment should be protected, and what must change by Monday morning.

The CFO asks who owns the economic trade-off. The manager says finance, operations, and sales will need to align. The COO asks what the plant, branch, project

or service team should do differently this week. The manager says the teams are working through the implications. The CHRO asks whether the manager has the authority to make the call. The answer becomes less clear.

Nobody in the room accuses him of incompetence. That is what makes the moment more severe. The manager has done what the system rewarded. He has kept stakeholders informed, avoided premature commitment and protected himself from being wrong alone. In many companies, that behaviour passes as maturity. Under AI, it may start looking like the absence of judgment.

This is why the debate on AI and jobs misses the first managerial disruption. AI may not replace managers in one dramatic sweep. It may separate managers who coordinate information from leaders who convert intelligence into accountable action. The first group will still sound busy. The second group will become visibly scarce.

The difference is not technical skill. A manager does not become valuable by competing with AI at analysis. The manager becomes valuable by asking what the analysis means for customers, people, risk, cash, delivery, and reputation. The manager becomes valuable by deciding when evidence is strong enough to act, even if it is not complete enough to feel safe.

Indian companies should take this seriously because many large organisations are coordination-heavy. Reviews, approvals, committees and escalation chains create the appearance of discipline. A manager who keeps everyone aligned is often considered effective, even when alignment delays the decision everyone is avoiding. AI will make that pattern harder to hide because it will reduce the time needed to know and increase the pressure to choose.

The uncomfortable truth is that some managers have been promoted for fluency, not judgment. They can speak the language of strategy, risk, customer focus and transformation. They can present complex situations without panic. But when the room asks what should be sacrificed, accelerated, stopped or defended, their authority thins out.

This does not mean companies should reward reckless decisiveness. Judgment is not speed alone. It is the disciplined ability to interpret incomplete evidence, weigh consequences, choose a course and accept responsibility for the outcome. It includes the courage to revise a decision when the facts change. It also includes the restraint to wait when waiting is the right call.

But waiting must itself become a decision, not a refuge. In an AI-enabled enterprise, the question will move faster from “what is happening?” to “what are you going to do about it?” Managers who answer that question with process language will look increasingly exposed.

The CEO now sees the review differently. The issue is no longer whether the manager has command of the facts. He does. The issue is whether the facts have produced judgment. The presentation proved visibility. It did not prove leadership.

That distinction should change how companies assess management capability. Do not ask only whether a manager gives accurate updates. Ask which trade-offs the manager can resolve. Do not ask only whether risks are visible. Ask what action follows when the risk crosses a threshold. Do not ask only whether teams are aligned. Ask whether alignment has produced a decision or merely protected everyone from one.

It should also change leadership development. Companies should not train managers only to use AI tools. They should train them to stand in the space AI will expose: the space between knowing and deciding. That requires sharper decision rights, clearer accountability, better problem framing and a culture that does not punish every well-reasoned call that later needs correction.

You entered the review impressed by the manager’s command of the facts. You leave with a different concern. The facts were never the full test. The real test began when someone had to decide what the facts demanded.

AI will expose managers not because it is hostile to them, but because it will remove the camouflage around managerial work. Reporting will no longer look like leadership. Coordination will no longer look like ownership. The next generation of managerial value will belong to those who can take intelligence into the room, face the trade-off and make the call.

(The author is Senior Director & Global Head – Design-to-Deliver, Digital Transformation & Industrial AI at Bosch Software and Digital Solutions. Views are personal.)