Vol.026 — AI Adoption Series: I Rolled Out AI. My Team Isn’t Using It. WHY?
Date: 2026-05-13 | Newsletter
Key Summary
Launching the AI Adoption Series, Zenta uses an anonymised ASEAN client — a mid-sized HR placement agency — to examine why rollouts stall even when the appointed team is willing. The counterintuitive finding: the two people leading the rollout are the ones slowing it down, not out of refusal but because the more they learn about what AI can actually do in their industry, the quieter they become. Zenta names the hierarchy of signals: a constructive NO (“I tried this, here’s where it broke”) is the most valuable thing a rollout team can produce; a silent YES — agreeing in meetings, not actually using the tool — is the trap; and silent NOTHING is the worst, because by the time it’s visible someone is being fired and the underlying cause is still unknown. In high power-distance ASEAN cultures, and especially in Japan, silent YES is the default. Carnegie Mellon’s most-discussed AI agent paper of the year reinforces the diagnosis: most AI agent failures are organisational design failures, not model failures. Functional AI Partners recommends three structural fixes: name the role-change explicitly before the rollout starts, reward constructive NOs out loud in every meeting, and replace “corporate ChatGPT plus webinars” with one specific task per person per week with a Friday report-back. Do it simple. Make the silence visible before it becomes a firing.