The strongest case for returning to the office is not control, surveillance, or nostalgia. It is institutional performance.

Organizations are not merely collections of individual tasks. They are systems of apprenticeship, trust, judgment, culture, and shared accountability. Remote work answered one important question: can many individuals perform outside the office? Yes.

But leaders now face a harder question: can the organization continue to build the capabilities it needs when too much of work becomes physically disconnected?

Remote work may preserve individual output while weakening the conditions that produce future organizational performance. That risk is harder to measure than commute time, employee satisfaction, or short-term productivity. But it is precisely the kind of risk leaders are paid to see.

The office should not return because leaders need an audience. It should return because organizations need shared context.

AI is making something visible that was easier to ignore before. As systems improve, organizations no longer struggle to see what’s happening. Performance signals surface earlier, patterns are clearer, and issues appear before they escalate. But visibility doesn’t resolve what follows. It simply moves the pressure to a different point—where information must become a decision.

That’s where variation begins to show. Two managers see the same signal and respond differently. Not because they lack capability, but because the decision itself isn’t consistently defined. Over time, those differences compound. What looks like a problem of management layers is often something else entirely—the absence of a clear, shared model for how decisions are made once the signal appears.