Leaders use "trust" and "psychological safety" as if they were the same thing. They are related, but they are not the same. And the gap between them explains one of the most confusing experiences in leadership: a team that genuinely trusts you and still won't tell you the truth. 

My team trusts me.

I hear this sentence from leaders more than almost any other. Usually, it arrives as a defense. We're talking about a quiet meeting, a problem that surfaced late, a resignation nobody saw coming, and the leader says some version of, "But my team trusts me. They know I have their back."

And most of the time, I believe them. 

They keep their word. They advocate for their people. They follow through. If you asked their team members privately, "Do you trust your manager?" the honest answer would be yes.

And the meetings are still quiet. The mistakes still surface late. The resignation still felt unexpected. That isn't hypocrisy on the team's part. It isn't a failure of the leader's character. It's what happens when a leader has built trust and assumed that safety came with it. It doesn't.

Two definitions worth getting right.

Trust has been studied for decades, and the definition that most researchers work with comes from Roger Mayes, James Davis, and David Schoorman. In their 1995 model, trust is:

"The willingness of a party to be vulnerable to the actions of another party based on the expectation that the other will perform a particular action important to the trustor."

Notice the shape of that sentence. There is a person. There is another person. There is a willingness to be vulnerable TO THEM, based on their expectation about what THEY will do. Mayer and his colleagues found that expectation rests on three judgments about the other person: their ability, their benevolence, and their integrity. Can they? Do they care about me? Will they do what they say?

Psychological safety comes from Amy Edmondson's 1999 research on hospital and manufacturing teams: 

"A shared belief held by members of a team that the team is safe for interpersonal risk-taking."

Both definitions involve vulnerability. That's why the two ideas often get confused. But look at where the vulnerability points. Trust is my willingness to be vulnerable TO YOU. Psychological safety is my belief that the GROUP will not punish me for being vulnerable. 

Edmondson herself drew the line in a 2004 chapter that deserves to be better known. She named three ways psychological safety differs from trust. I'll add two more that show up in practice. 

Difference 1: Trust is about them. Safety is about you.

When I decide whether to trust you, I'm evaluating YOU. Your competence. Your intentions. Your history. 

When I decide whether it's safe to speak, I'm evaluating my own exposure. If I ask this question, will I look ignorant? If I admit this mistake, will I look incompetent? If I push back, will I look difficult?

Trust looks outward. Safety looks in the mirror. 

That's why a leader can be entirely trustworthy and still preside over a silent room. The team isn't running a calculation about the leader's character. They're running a calculation about their own risk. 

Difference 2: Trust is about the future. Safety is about right now.

Trust is an expectation about what someone will do over time. Will she follow through on the promotion conversation? Will he protect the team when budgets get cut? It's forward-looking, and it accumulates. 

Psychological safety is about the immediate consequences of a specific act. It's decided in the half-second before someone opens their mouth. Not "will this person come through for me over the next year," but "what happens to me in this room, in the next thirty seconds, if I say this?"

Trust has a memory. Safety has a clock. A leader can spend two years building a track record and lose the room's safety in a single meeting with one sigh, one "we've already decided," one visible flash of irritation at a question. The trust survives. The safety doesn't. 

Difference 3: Trust lives between two people. Safety belongs to the whole group. 

This is the difference that matters most for leaders, and the one most often missed. 

Trust is fundamentally dyadic. I trust you. You trust me. It's one relationship at a time, and it can be different for every pair of people on a team. 

Psychological safety, in Edmondson's research, is a group-level property. It's a shared belief about the climate. It emerges from what everyone has watched happen to everyone else. 

So let's consider what this means. A team member can trust their manager completely and still fear their peers. They can believe the leader will handle a hard truth well and still remember what a colleague did to the last person who admitted confusion. The eye-roll from across the table. The joke landed a little too hard. The idea was quietly dismissed and never came back.

The leader didn't do any of that. The leader may not even have noticed. But the room learned. Which means that a leader can't build psychological safety through one-to-one relationships alone. Trust is built in private. Safety is built in public. 

Difference 4: Trust is built slowly. Safety is built in moments.

Because trust rests on ability, benevolence, and integrity, it's earned the way reputations are earned: through consistency over time. Keep your word. Follow through. Show up. There's no shortcut, and that's appropriate. 

Safety is built differently. It's built in what Edmondson calls the response to interpersonal risk. Someone asks a question that might be obvious. Someone admits they're behind. Someone disagrees with the plan. What happens in the next three seconds is where safety is won or lost. 

This is actually good news for leaders. You can't build trust with a new team in a week. But you can begin building safety in your first meeting, by how you respond to the first person who takes a risk. 

Difference 5: Trust breaks when you fail someone. Safety breaks when anyone is punished for speaking. 

Trust is damaged by betrayal, broken promises, incompetence, or self-interest. Those are personal failures, and they can be repaired personally. 

Safety is damaged by consequences, and the consequences don't have to come from the leader. A colleague's sarcasm. A sigh. A "that's not how we do things here." The silence after someone shares a half-formed idea. 

Nobody has to betray anyone. The room simply learns that speaking has a cost. 

Four rooms.

 Put the two dimensions together and every team lands in one of four rooms.

High trust, low safety: the loyal, quiet room. People trust the leader. They'd follow her anywhere. And they still won't risk looking wrong in front of each other. This is where most leaders who say "my team trusts me" are actually standing. The meetings are pleasant. The information is incomplete. 

Low trust, low safety: the guarded room. Nobody trusts anybody, and nobody takes risks. Silence is survival. This room is at least honest about what it is. 

Low trust, high safety: the polite room. Team norms make it safe to speak, but people don't believe the leader will do anything with what they say. So they contribute, but only about what's cheap to contribute. Ideas, yes. Bad news, rarely. 

HIgh trust, high safety: the learning room. People believe the leader will come through and they believe the room won't punish them for their candor. Mistakes surface early. Disagreement is normal. This is the only room that learns faster than it makes mistakes. 

The point of the grid isn't to label your team. It's to notice that the top-left room exists at all. Trust without safety is not a contradiction. It's common.

Why this matters for how you lead. 

If you assume trust and safety are the same thing, you will make three predictable mistakes.

You'll measure the wrong thing. You'll ask, "Does my team trust me?" and when the answer is yes, you'll stop investigating it. The better question is: "What happens to people here when they're wrong out loud?"

You'll build in the wrong place. You'll invest in one-to-one relationships, which is good, and expect that to change the meeting, which it won't. Safety is a property of the room. It has to be built wher ethe room can see it. 

You'll miss the peer dimension entirely. You'll monitor your own reactions and ignore what team emmbers do to each other. But safety is shared. One colleague who makes people pay for speaking can undo months of a leader's good work, and the leader may never see it happen. 

Here's what to do instead.

Here's the practical translation:

Keep building trust the slow way. Keep your word. Follow through on what people tell you. Be honest about what you can and can't change. This is necessary. It just isn't sufficient. 

Build safety out loud. When someone takes an interpersonal risk in front of the group, treat your response as the most important thing you'll do that day. "Thank you for raising that." "Say more." "I hadn't thought of it that way." Where other people can hear you. Because everyone else is deciding whether to go next. 

Protect the room, not just the relationship. If a team member's reaction makes speaking expensive for someone else, that is your business. Address it, privately and promptly. The team is watching whether you will. 

Ask the room, not the individual. Instead of "Do you trust me?," try: "What would be hardest to say in this meeting?" or "What happened the last time someone here was wrong out loud?" The answers will tell you which room you're in.

Trust is necessary. It is not sufficient. 

The leaders I work with are, by and large, trustworthy people. They keep their word. They genuinely care. Their teams know it. 

That's why the quiet meeting confuses them so much. 

But your people aren't measuring you. They're measuring the room. They keep their word. They care. Their teams know it.

Trust is a bet on a person. Safety is a read on the room. 

Build both. Just don't assume the first one gave you the second.

References

Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2).

Edmondson, A. C. (2004). Psychological safety, trust, and learning in organizations: A group-level lens. In Kramer & Cook (Eds.), Trust and Distrust in Organizations.

Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995).  An integrative model of organizational trust. Academy of Management Review, 20(3).