Team Pulse by CoastZone
All Episodes

Trust Cracks, AI Swarms, and the Future Team

This episode examines how political and cultural differences can quietly erode team trust, and why leaders need to avoid the hidden traps of certainty, inconsistency, emotional reactivity, and self-justification. It also looks ahead to the rise of AI agent swarms, the reality of low AI adoption, and the challenge exhausted managers face when asked to lead both people and machines.


Chapter 1

The Friction of Difference and the Cracks in Team Trust

Isabella

It, it, it actually blew my mind when I looked at the, the 2026 Edelman Trust Barometer. We talk so much about trust, right? But get this, 34% of people surveyed said that if their team leader held different political beliefs, they would actually put less effort, like, less physical effort, into helping that leader succeed.

Jesper

Oh, wow. 34%. That is, uh, that is a massive, silent tax on a team before you even, you know, before you even hit a single project milestone or, or face a real crisis. People are just, um, quietly pulling back because of value sorting.

Isabella

Exactly! It is value sorting. We are sorting ourselves into groups and, and the scary part is, the same study says 78% of employees trust their employer to do the right thing. That is the highest of any entity. So we, we trust the big company, but we do not trust each other across the desk. 70% of people said they are hesitant to trust someone who differs from them in values or, or culture. And 42% would actually rather change departments than report to a manager with different values.

Jesper

Yeah, that, that really tells us that the team is, is becoming this, uh, this battleground, or, or maybe the last place where we actually have to interact with people who do not think like us. But as leaders, we, we often completely waste that difference because we do not know how to handle the friction. There was this, uh, July 15, 2026 article in Harvard Business Review by Susan MacKenty Brady, Stuart Kliman, and Leslie Smith. They literally wrote a book called All the Difference, and they talk about these four hidden traps of team dynamics that quietly ruin everything.

Isabella

Okay, what are they? Because I feel like I, I probably fall into at least two of them on a bad Monday.

Jesper

Well, the first one is Certainty. Just, you know, being so convinced you are right that you completely stop listening to anyone else. Then there is Inconsistency between words and actions. Like, saying you want people to speak up, but then, uh, you know, punishing them or going totally quiet on them when they actually do it. And then we have Emotional Reactivity, letting the gut reaction speak first, and Self-justification, where you just defend your own position instead of, you know, actually looking at the facts. We really need to, uh, være ops på how easily we slide into these when the pressure is on.

Isabella

Oh, the inconsistency one is so real, especially in Denmark. We have this, this beautiful low power distance and we love our consensus, right? We always say we want everyone's input. But then, if someone actually disagrees, we, we tend to smooth it over immediately. We reach for en løsning alle kan leve med, a solution everyone can live with, way too quickly. We don't let the disagreement do its work. We, we just want the cozy harmony back.

Jesper

Yes! It is, uh, it is exactly that. We, we run away from the friction because it is uncomfortable. But, you know, the Nordic sweet spot in conflict is actually being hard on the issue, but warm on the person. You can disagree intensely about the data, but you still, uh, you still care about the colleague. If we do not do that, we are just smoothing away the very diversity we hired people for. The leader's behavior becomes the absolute ceiling for what the team can handle.

Isabella

It really does. I, I think about my own behavior. I, I definitely fall into certainty. I get an idea and I am like, yes, this is the way, and I have to actively force myself to ask, wait, what else could be true here? Which is actually one of the six actions they recommend in the book, right? Cultivating self-awareness, returning to respect.

Jesper

Yes, yes, it is! And, and if we can just pick one of those actions, you know, communicate with candor, or, or seek the full story before we jump to conclusions. Yes! That is how we start to lift that ceiling before our next tense meeting.

Chapter 2

Swarms of AI Agents and the High-Demand Danish Sweet Spot

Isabella

And speaking of difference, the next teammate joining the team might not even be human. I was looking at McKinsey's work on the agentic organization. They describe this wild scenario where a small human team of, say, two to five people is supervising a swarm of 50 to 100 specialized AI agents. Like, an agent factory running end to end processes. We move above the loop, steering and validating rather than doing the manual work. We get M-shaped supervisors who are generalists, and T-shaped experts who handle the weird exceptions. It sounds like a dream for flat, high trust Danish teams, doesn't it?

Jesper

It, it sounds great on paper, Isabella, it really does. But, uh, let us look at the reality check from Gallup. Their 2026 State of the Global Workplace data says that only 12% of workers strongly believe AI has actually changed how they work. Only 12%! And the biggest factor is not the technology itself. Gallup's CEO, Jon Clifton, said, and I love this quote, "Even the most sophisticated neural network cannot overcome an indifferent team leader." If the manager does not actively champion it, the team simply will not use it.

Isabella

Whoa. 12% is, is tiny compared to the hype. And, and if the manager is the bottleneck, that is a huge problem because Danish leaders are notoriously slow at adopting AI compared to their employees. We have this gap where the staff is experimenting, but the managers are, are kind of lagging behind.

Jesper

Exactly. And the managers themselves are completely depleted. Gallup's data shows manager engagement fell from 31% in 2022 to just 22% in 2025. They are exhausted. So now we are asking these burnt out managers to not only lead people through trust cracks, but also to learn how to orchestrate a swarm of 50 AI agents? It is, uh, it is a lot to ask.

Isabella

It is. Though, I did see that Aarhus University is trialling a Danish AI tool called Lederweb to help leaders lead more empathetically. So maybe the AI can help us manage the humans, while we manage the AI? But, but it brings us back to what leadership actually is. We have this fresh PhD study from May 2026 by Anne Malene Jensen at Aalborg University. And she found something that challenges the whole Danish concept of tryghed, of psychological safety.

Jesper

Oh? Tell me. What did she find?

Isabella

Well, she basically says psychological safety works best when it is paired with highly ambitious demands. She said, and I quote, "Målsætningerne giver retning, men fællesskabet giver ressourcerne til at nå dem." Goals give direction, but the community, the safety, gives you the resources to reach them. If you have safety without demands, the team just coasts. It becomes a cozy, warm, low-performing club. But if you have demands without safety, you get fear. You need both.

Jesper

Hmm. That is, uh, that is highly pragmatic. It is like Amy Edmondson's learning zone, but with a, a very clear Danish focus on ambition as the activator. We often mistake safety for just, you know, being nice and comfortable. But real safety is what makes high-pressure, ambitious work survivable.

Isabella

Yes! And if you pair that with the MIT Sloan finding from this year, that consensus is actually built by social connectors in the team, not the high-status people on the org chart, then the task for the leader is clear. You set the high bar, you build the safety, and you route the change through those key connectors.

Jesper

But it still leaves us with this, uh, this massive tension, doesn't it? If our managers are already so depleted, with engagement down to 22%, how on earth are they supposed to hold these high-demand, high-safety environments while steering a hundred AI agents? It is, uh, it is a bit of a paradox. We are asking them to be superhuman exactly when they feel most humanly exhausted. But, uh, I guess we will have to find that balance, otherwise the system just breaks.