AI is easing your team's burnout. It may also be quietly hollowing out the team.

Your team's numbers are up and the stress is down, and you should feel good about that. Mostly you do. But walk the floor and it is quieter than it was. People ask the AI the thing they used to lean over and ask the person beside them. The burnout has eased, which is real and worth having. And something else, harder to name, has thinned along with it. You can feel the trade before you can measure it.

Here is the trade, named plainly. The same tool that is lifting your team's output is also quietly standing in for the small human moments that used to hold the team together. That is not a reason to slow the productivity. It is a reason to notice what it can cost if you leave it to chance, and to design for the part that does not automate.

Is AI really weakening connection at work?

The data suggests it can, and the same data shows why it is easy to miss. Workday, the workforce software company, surveyed 2,150 full-time employees across seven countries in 2026. The headline is good news: 86% said they feel more productive because of AI, and much of the burnout it eases is real. But underneath sits a quieter finding. 37% now use AI for companionship, and one in five Gen Z workers said they took time off this year because of loneliness or isolation.

Put those together and the shape is clear. The tool is doing more than tasks. For a growing share of people, it is filling a social gap that a colleague used to fill. Gallup's workplace research adds the context: engagement and belonging have been under strain for years, and they are what drive discretionary effort and retention. So this lands on ground that was already soft. It is a close cousin of the reason the human side of work is the part you cannot automate your way past.

Why does this count as a business cost, not a soft one?

Because belonging is not a perk. It is a driver of the things you already measure. Engaged, connected teams keep their people longer, give more discretionary effort, and share what they know. When connection thins, those quietly weaken, and they weaken in ways that do not show up on the productivity dashboard for a while. So a leader can post a strong efficiency number and be eroding the base underneath it at the same time. The gain is visible and immediate. The cost is real and delayed, which is exactly what makes it easy to miss.

The lower-altitude question is which AI tools to roll out, and how fast. Fair enough. The higher one is what happens to your people as the tools absorb more of the day. A productivity win that quietly costs you cohesion is not a clean win. It is a trade you made without pricing one side of it, and it is the same blind spot behind so many AI rollouts that hit their targets and lose their people.

So how do you keep the team human as AI scales?

You design for connection as deliberately as you design for efficiency, rather than assuming it survives on its own. The aim is to let AI take the drudgery and use the time it frees to strengthen the human side, not vacate it. Work it in this order:

1. Point AI at the drudgery, on purpose. Aim the tool at the repetitive load so it gives people time and energy back, rather than letting it quietly replace their contact with each other.
2. Protect the human rituals. Keep the shared lunch, the standup, the desk-side question alive by design. These are the small moments a team is actually built from.
3. Watch the quiet ones. The cost concentrates in the people leaning hardest on the tool for company. Notice who has gone quiet, and reach for them as a person.
4. Spend the freed time on people. When AI hands back hours, deliberately invest some in connection, mentoring and shared work, not only in more output.
5. Measure belonging, not just productivity. Track how connected and supported people feel, and treat a falling number as the early warning it is.

What could this look like when you get it right?

A team that is more productive and more connected, not one traded for the other. AI carries the repetitive load, the burnout stays down, and the hours it hands back go partly into the human side on purpose: time to talk, to teach, to solve things together. People still turn to the tool for the task and to each other for the rest. The efficiency is real, and the team it runs through is stronger for it. That is a business that has grown into AI rather than been quietly hollowed by it.

Picture the floor a year on. It is busy with human conversation again, not because the AI did less, but because you spent what it gave back on the things that make a team a team. Your quietest people are talking to colleagues, not only to a screen. You took the productivity and kept the belonging, because you designed for both. That capacity to keep people connected through change is the most human part of leading, and it is where lasting advantage now sits.

About the author

Thomas Green

British technology futurist, AI keynote speaker and advisor. Thirty years across enterprise technology and AI strategy, helping leaders navigate the future of work. The futurist who died.