Three of your collaborators’ biggest AI fears are already management problems

“I’m using AI every day. I’m more efficient. But I’m less sure that the conclusions I’m reaching are actually mine. And I don’t know what to do about that.”

That sentence came from a real conversation. And in some version, it’s happening inside your organization right now.

The four most commonly expressed concerns of AI use are: unreliability (26.7%), job and economic anxiety (22.3%), loss of autonomy and agency (21.9%), and cognitive atrophy (16.3%). Source: Anthropic

The numbers are easy to read flat. But the study found something that changes the picture: respondents raised an average of 2.3 concerns each. These fears overlap and reinforce each other. The person worried about unreliability is often the same person worried about their job, worried about their thinking, worried about whether the conclusions they’re reaching are still theirs.

And sitting alongside all of it: hope. Excitement about what AI makes possible. Both things, in the same person, at the same time.

That complexity is what’s actually inside your organization right now.

And three of those four fears are already management problems.
‍

Unreliability, 26.7%

When people don’t trust AI outputs, they either over-verify everything, slowing down every workflow, or they stop questioning at all. (the second one is way more common than organizations want to admit, and way more dangerous.) Both responses produce the same outcome: your AI investment delivers less than it should, and decision quality degrades.

The real question unreliability raises is whether your people have the judgment to know when the tool is wrong. That’s a human capability question. And it’s one most organizations haven’t asked yet.

Autonomy and agency, 21.9%

The concern here isn’t that people are using AI too much. It’s that they’re using it passively. There’s a difference between collaborating with a tool and outsourcing your thinking to it. Between delegating a task and surrendering a decision.

When that line blurs, and it blurs easily because AI is designed to feel helpful, ownership erodes. People produce more output while feeling less responsible for it. Accountability becomes abstract. And the organization loses something it didn’t know it was losing.

Cognitive atrophy, 16.3%

This one is the slowest and the most expensive. In Anthropic’s study, 46% of those who expressed concern about cognitive atrophy said they had already seen it firsthand. Not a theoretical fear. A reported experience.

The ability to check, challenge, and contextualize the judgment that makes your organization valuable, weakens when it isn’t practiced. AI can increase immediate performance while degrading the capability required to evaluate that performance. Everything you don’t sustain, you lose.

‍

The fourth concern, jobs and the economy, is different.

No workshop solves job insecurity. But Anthropic’s data found something worth looking at: people experiencing the greatest AI productivity gains were also more worried about job displacement. Turns out, demonstrating efficiency doesn’t reassure employees. Sometimes it increases the perceived threat (keep that in mind before the next all-hands, honestly).

What changes the experience is orientation. When people understand what the organization is trying to build with AI, what remains uncertain, and what role human judgment plays in that future, the fear doesn’t disappear, but it stops being paralyzing. They move from “AI transformation is happening to me” to “I understand what we’re trying to become, and I have some agency in getting there.”

That requires transparent leadership. And it requires people who trust their own capacity to adapt.

‍

Companies have spent the last two years teaching people how to use AI.

The next question is whether their people know how to think while using it.

When people learn to think when AI thinks with them (and not for them!), they reclaim ownership of their thinking, of their work, of their conclusions. And they become the kind of people who can participate responsibly in what we’re collectively building right now.

And this plays a big role in whether your AI investment compounds or stalls.

It’s with that in mind that I created,

Own Your Thinking, a one-day intensive for teams actively rolling out AI.

Built around three of those four most commonly expressed concerns. And one core method most organizations have never invested in: the relationship people have with their own mind, how they decide, how they evaluate, how they stay the author of their own conclusions.

Know more about Own Your Thinking solution