The CEO is bought in. The frontline is using the tools. The initiative is stalling anyway.
There is a pattern I have seen often enough that I think it deserves a name. Executive sponsorship is real — the CEO talks about AI on every earnings call and means it. The frontline adoption is real too — the people doing the work have the tools and are using them. By every metric leadership tracks, this should be working.
And it is not moving. The friction is coming from the layer in between.
The directors and the VPs — the middle management layer whose entire job is to translate strategy into execution — never got a straight answer to the one question they actually have. Not “is AI good.” They know it is good; they read the same reports. The question is: what does this mean for me, my team, my headcount, my mandate, and the way my performance gets judged next year? And nobody answered it, because everybody assumed somebody else had.
That silence does not stay silent. It converts. It becomes hesitation on the resourcing decisions only that layer can make. It becomes projects that are technically approved and somehow never quite staffed. It becomes a kind of passive resistance that no one can point to directly, because no one is refusing anything — they are just not leaning in, and leaning in was the whole plan.
The easy read is that this is politics. Middle managers protecting turf, slow-walking the thing that threatens them. I do not think that read is fair, and more importantly I do not think it is useful. Turf is a symptom. The disease is ambiguity — you handed a layer of the organization a mandate that quietly reshapes their role, and then you never told them what their role becomes on the other side of it.
These transformations do not collapse because the frontline can’t learn the tools. They stall because the middle layer was handed a change that affects them more than anyone and briefed on it less than anyone. It is an organizational design problem wearing the costume of a political one.
The fix is not a town hall. It is answering the actual question, per role, in specifics: here is what your job becomes, here is what you are now accountable for, here is what stops being your problem, and here is how we will judge whether you did it well. Uncomfortable, concrete, and early. The organizations I have seen get through this are the ones willing to have that conversation before the resistance shows up, not after they have spent six months diagnosing it as something else.
Worth asking about your own program: has anyone actually told the director-and-VP layer what their job becomes — or are they still filling that silence in on their own?