Your AI Strategy Is Probably Breaking at the Manager Level

The newest HBR ideas keep circling one uncomfortable reality: executives may be excited about AI, but the people expected to make it work on the ground are often overloaded, unconvinced, or simply out of room.

A lot of AI strategy still gets discussed from the penthouse. The executive team sees cost savings, speed, competitive positioning, and a cleaner future. The manager sees another system to roll out, another workflow to fix, another set of half-clear expectations, and another wave of pressure dropped directly into the middle of an already crowded week.

That gap is not small. Harvard Business Review recently put it plainly in a piece arguing that managers and executives increasingly disagree on AI, and that the disagreement is costing companies. HBR followed with another piece urging leaders to close the gap between AI ambition and execution.

That is a useful frame because it explains why so many AI rollouts feel impressive in the announcement phase and clumsy in the living phase. Leaders often experience AI as a strategic story. Managers experience it as operational weather. One group is discussing transformation. The other is trying to figure out whether the tool actually works, whether the team trusts it, who owns the process, and what happens when the software gets it wrong. HBR’s recent focus suggests this is becoming one of the core management problems of the AI era.

The problem gets sharper when you look at the broader workplace data. Gallup’s 2026 State of the Global Workplace says global employee engagement fell again in 2025, and it highlights managers as the group where the decline has been most acute. Gallup reports that global manager engagement dropped from 31% in 2022 to 22% in 2025, with the biggest one-year drop occurring from 27% in 2024 to 22% in 2025. It also estimates that low engagement cost the world economy about $10 trillion in lost productivity last year.

That chart above matters because it shows the part of the workforce companies keep leaning on while quietly draining it. If managers are the transmission layer between strategy and execution, and that layer is becoming less engaged, then companies should not be surprised when rollout quality gets shaky. Gallup goes even further and says managers are critical for AI adoption, noting that in a separate U.S. workforce survey from the first quarter of 2026, one of the strongest predictors of whether employees actually use AI is whether their direct manager actively champions it.

This is where many businesses get the sequence wrong. They buy the tools, announce the initiative, train the executives, and then assume the rest will trickle down. But AI does not trickle well. It has to be translated into workflows, permissions, norms, judgment calls, follow-up habits, and real-world behavior. That translation job usually lands on managers, and HBR’s recent writing implies that too many companies are treating those managers as if they are neutral delivery pipes rather than overloaded people with their own skepticism, constraints, and fatigue.

There is also a trust problem hiding in here. Executives may talk about AI in terms of upside, but managers are often the ones who have to absorb the downside first. They deal with employee anxiety, output quality issues, process confusion, and unclear accountability. They are the first to hear “this tool slows me down,” “I do not trust this answer,” or “are we being measured on this now?” If leadership is not hearing those concerns directly, it can mistake silence for alignment. That is usually where strategy starts getting dressed up as success before it has earned the title.

The companies that handle this better will likely do a few unglamorous things. They will diagnose manager sentiment before calling the rollout a win. They will simplify expectations instead of stacking new AI responsibilities on top of old work. They will train managers as interpreters, not just users. And they will stop assuming that executive enthusiasm means operational readiness. HBR’s guidance to close the gap between ambition and execution sounds simple, but in practice it is a warning against one of the oldest habits in business: confusing a top-level decision with a company-wide capability.

For founders and smaller teams, the lesson is even more practical. If your business is small enough that you do not have layers of management yet, then good. Learn from this before you create the problem. Do not roll out AI in a way that leaves the people closest to customers, process, and delivery carrying all the ambiguity. Put clear use cases in place. Be honest about what the tools can and cannot do. Make someone explicitly responsible for the workflow. And remember that the real question is not whether leadership is impressed. It is whether the people doing the work can actually use the tool without their job turning into a longer, stranger version of itself.

That is what the best recent HBR thinking seems to be nudging leaders toward. AI is not just a technology story. It is a management story. And if the managers are exhausted, unconvinced, or disengaged, then even the smartest AI strategy may end up sitting in the company like expensive furniture nobody really uses.