AI as a diagnostic for dependencies

In my previous post, I explored what happens when AI exposes where decision authority sits (or doesn't). But even when decision authority becomes clearer, progress does not automatically accelerate. Work still has to move between teams, and the moment it crosses a boundary, dependencies come into view.

In many organisations, dependencies are familiar but rarely examined. Teams know who they rely on and roughly how long things tend to take. Delays are anticipated, hand-offs are managed through relationships. Over time, this coordination becomes part of the normal cost of getting things done.

The difficulty is not that dependencies exist. It is that their impact is often underestimated. When work begins to move more fluidly, whether through automation, improved tooling, or simply greater autonomy, it reaches team boundaries more frequently. What once felt like occasional coordination now becomes a recurring negotiation. Context needs to be re-established, assumptions need to be aligned, priorities need to be reconciled.

What becomes visible at that point is not just the presence of dependencies, but the degree of coupling in the system. Teams that seemed largely independent turn out to be synchronised in subtle ways. A change in one area requires adjustments elsewhere, shared platforms absorb requests from multiple directions, coordination overhead increases even when individual teams are operating efficiently.

None of this reflects poor performance; it reflects design. In slower environments, coupling is masked by experience and goodwill. People bridge gaps manually and absorb the cost. Under AI-driven pressure, that absorption capacity diminishes. Dependencies do not disappear with better tools- they become more apparent.

For leaders, perhaps the important question is not what dependencies exist, but whether they are intentional. If you mapped how work actually moves from idea to outcome, would the hand-offs reflect deliberate design, or historical layering?

A simple starting point is this: trace one recent initiative end-to-end and note every team it required to progress. If the number surprises you, or if progress depended heavily on negotiation rather than clarity, the signal is already there