The Human Layer

Most organizations now use AI, yet few capture the value they expected. The gap rarely lies in the model or platform; it stems from the conditions around the technology: how leaders lead, how work and roles are designed, how decisions are governed, and whether the technical foundation supports the ambition.

The budget was approved. The platform was selected. The business case was compelling, the pilot was promising, and the rollout was declared a success. Twelve months later, something is quietly wrong.

Decisions that were supposed to be faster are still bottlenecking at the top. Teams that were supposed to be empowered are still waiting for direction. Leaders who championed the initiative in the boardroom are reverting, in practice, to the behaviors and instincts that predate every tool they were handed. AI adoption is real on paper but limited in practice.

When adoption stalls, the response is predictable: more training, better communications, a change workstream, a town hall where executives demonstrate the tools, and sometimes a technology review that recommends adjustments to the implementation plan.

Each of these has a place. Each also tends to address one layer of a problem that spans several. Training assumes the gap is informational. A technology review assumes it is architectural. A change workstream often assumes it is a matter of communication. In our experience, stalled adoption usually reflects all three at once, together with something harder to see: leaders who have not yet changed how they lead, roles that were never redesigned around the new capability, and decision rights that no one clarified.