The most consistent thing I’m hearing from leaders about AI is concern. They’re unsure how to use it, what it’ll cost, or where to start. But mostly they’re worried their peers in other organizations have already figured it out and they haven’t.
The pressure is real, and most of the advice on offer makes it worse, because nearly all of it says the same thing, move now or fall behind. I want to make a different case. The place to look first isn’t AI, it’s your managers.
Start with your managers
Your managers are under a pressure that’s been building for years. The average manager today has more direct reports than a manager did a decade ago, and the layers that used to sit between them and the work have been thinned out in the name of efficiency. They’ve become the place where every new initiative lands.
Last month I wrote about what that does to people over time. Capable people stop reaching. From the outside it can look like apathy. Up close it’s likely a calculation: they’ve concluded that offering more changes nothing, so they stop offering it. The manager layer feels this first and feels it hardest. What I didn’t do last month was answer the question of what to do about it. I want to share those thoughts now.
AI multiplies what’s already there
What I see most often is this. When a leader recognizes that their managers are stretched past what’s reasonable, AI looks like the lever. The promise is that it lifts the load, clears the admin, hands people their time back. That’s an understandable hope, but based on current research it won’t deliver the result you’re after.
AI sits on top of the way your organization currently works. Where that work is fragmented, AI amplifies the fragmentation rather than fixing it. Harvard Business Review has a name for one version of this: “workslop.” AI produces work quickly, but the quality isn’t assured. Facts may be inaccurate, and the critical judgement that comes from human experience is missing, so additional effort is needed on top of the original task. Research by BetterUp and Stanford put that at close to two hours of rework each time. Much of the fixing is an additional burden that lands on the manager’s desk.
Waiting is the wiser move
For most organizations unsure of their next step, the wiser move is to hold off on large-scale AI investment until the landscape is clearer. The tools change every few months, and the approach that turns out to be right may not exist yet. It’s worth knowing that most organizations spending heavily on AI report no meaningful return on it so far, which suggests the rush to commit isn’t paying off the way it promised. It’s also worth asking what outcome you actually want, and what would deliver it. If that’s not easy to articulate, then reaching for AI isn’t the right move yet.
Where to look before you spend
Holding off on a big AI commitment doesn’t mean doing nothing. It means putting your effort where it holds its value whatever AI becomes, and that’s closer to hand than another platform. A lot of what’s wearing your managers down has nothing to do with technology. It comes from the way the work around them is set up, and most of it can be improved without buying anything.
You’ll know your own organization better than I do, so take what fits and leave the rest. A few things tend to be worth a look.
One is the meetings. Most teams have a few that recur because they always have, not because anyone decided they still earn the time. Another is communication. In a lot of organizations nobody ever agreed which channel is for what, so everything goes everywhere and people are interrupted all day. Decisions are a third. When a decision gets made in a meeting or a thread and then lives only there, the next person who needs it has to interrupt someone to ask what was agreed. None of these show up as the problem. They show up as normal, which is exactly why they’re easy to miss.
Then there are the tools you’re already paying for. Most organizations get a fraction of the value from what’s already on the desk. Of the AI seats companies have bought, roughly two-thirds go unused. Before spending on anything new, it’s worth asking what you’d get from using what you already have well. For a lot of leaders, that’s where the nearest return is.
The part that’s yours to remove
Some of this your managers can take on themselves. Some of it they can’t, and it helps to be clear about why. A standing meeting, a norm that runs across the whole team, the agreed place decisions are meant to live, these aren’t things one manager can reset alone. They were set above them, and that’s where they get unset. That part is yours. You’re the one who can take a recurring meeting off the calendar, or signal that it’s fine to question a long-running habit. Give your managers that room and they’ll do a great deal with it.
None of this depends on the AI question being settled, and the managers carrying your organization can’t wait for it to be settled either. The worry you have that your peers have figured out AI is misplaced; most are as unsure as you are. There’s no race to win here. There’s just the slower work of clearing what’s in your managers’ way, and that’s yours to start whenever you decide.