The AI Communication Gap
The tools showed up. The communication didn't.
44% of workers say their employer has no clear AI policy, or they aren’t sure one exists.
I’m normally not one to have a corporate policy for every conceivable issue, but AI is important. It has, and will continue to change the workforce and every employee is nervous about the potential elimination of their job.
Employees deserve to know what you think about AI and where you’d like it to go.
The scariest part of change is the unknown. If we don’t know, we automatically assume the worst.
This isn’t happening the way you’d think
Most of the AI coverage out there assumes a leadership team on a zoom call announcing a big rollout — new tool, new strategy, new mandate. That’s not what I’m seeing.
What I’m actually seeing is quieter and potentially messier: AI tools showing up randomly department by department, often person by person, with no real strategy behind them and no communication to go with it. Someone in marketing starts using it. Someone in ops finds a use case. Nobody sends an e-mail. Nobody writes a policy. It just... happens.
That absence of a plan may like nothing to a leader who isn’t thinking about it. To the employee sitting next to the tool, it isn’t nothing. It’s a vacuum, and people fill vacuums with their own conclusions.
The gap the data shows
89% of workers have already used AI for work. That adoption happened fast, and mostly on its own, without leadership setting the terms.
Meanwhile, one 2026 study found 78% of employers view AI positively, compared to just 51% of employees — a 27-point gap. A quarter of employees name trust in AI as a real concern. Only 12% of employers see it the same way.
And in workplaces where AI use comes with some restriction but no real explanation for it, over half of employees say they trust AI-assisted work less than they did before those restrictions showed up. Restriction without reasoning doesn’t read as caution. It reads as leadership hiding something.
Why this breaks trust
With no strategy and no communication, employees are left to build their own theory of what’s happening.
Will this replace part of my job (or all of my job)? Is my work being judged by a standard nobody explained to me? Is my manager using this tool on my performance reviews? Am I supposed to hide the fact that I use it, or is it fine?
Nobody’s answering these questions — not because leadership is stonewalling, but because nobody’s decided the answers yet. That’s the real gap. It’s not a communication failure in the traditional sense. It’s an absence of decision-making that gets read as one.
This is the same pattern I see with change in general: the thing that erodes trust isn’t the change itself. It’s being left to guess about it. AI just raises the stakes higher than most changes, because this one touches whether people’s jobs still look the same a year from now.
The central focus of leadership needs to be communication. Not compensation, not perks, not a culture deck. Most of what gets labeled a trust problem, an engagement problem, or a morale problem is really a communication problem wearing a different name. AI is the newest test of it.
What actually closes the gap
We’ve done thousands of employee surveys over the years, across dozens of organizations. I’ve never seen a single response where an employee said, “back off on the communication.” Not once. Your goal in communication is to actually reach that point. It never happens. That’s the bar. Here’s how you get closer to it with AI specifically:
Decide before you deploy — even informally. You don’t need a 12-page governance document. You need someone in the room asking “what are we actually trying to solve, and what does this mean for the people doing the work” before the tool spreads on its own.
Write down the policy you probably don’t have. Given that 44% of workers can’t tell you whether one exists, even a one-page guideline on what’s expected, what’s allowed, and what needs disclosure puts you ahead of the game.
It’s OK to say you don’t know. If you haven’t decided whether AI use will affect headcount or role scope, say that directly. “We don’t know yet” builds more trust than silence ever will.
Normalize disclosure early. If nobody’s ever said it’s okay to mention AI was used on a piece of work, most employees will assume it isn’t.
Ask what’s already happening. Before you write a policy, find out what tools your people have already adopted on their own. You likely have more AI use in your organization right now than you think — most of it undocumented, some of it worth building on.
The stakes
Unmanaged AI adoption doesn’t produce a workforce that pushes back. It produces one that works around the tool quietly, hides its use, and stops trusting leadership to have a plan — for this or for whatever comes next.
The gap doesn’t close by making a bigger announcement. It closes by actually deciding something, communicating it plainly, and admitting what you haven’t figured out yet.
This was never really an AI problem. It’s a communication problem. And the solution is to…communicate.