Most AI adoption is going better in meetings than in practice.
In status updates, people nod. They ask thoughtful questions. They give positive-sounding updates. The signal leadership reads is: this is going fine.
What's actually happening below that surface is different. People carry on doing the work the old way, find ways around the new AI tool, find reasons the new workflow doesn't quite fit. WRITER's 2026 enterprise AI adoption survey put a number on it: 29 percent of employees admit to actively working around their company's AI strategy. 44 percent among Gen Z. Those are the people willing to say it out loud.
Look, people are busy. They have targets to hit. They default to ways of working they know produce results, and a new tool sitting on top of that hasn't yet earned a place in how they actually work. Adopting it costs them time today for a benefit that's promised tomorrow. Rationally, they wait. None of this is sabotage. It's inertia, and it stalls your rollout.
And this inertia isn't fully conscious. Most people don't know what their job looks like on the other side of the AI rollout. They suspect the answer isn't reassuring. Or that it's been decided without them. So they quietly keep their existing value visible in the old way, while waiting to see how things land.
The fix for this isn't more training. Training answers the question "how do I use this tool?" The question people actually have is "what does my job look like after this?" Those are not the same question.
Showing people the after-picture is what matters. Show them the bits of their role that AI will quietly absorb: the data entry, the formatting, the repetitive admin nobody got into the work to do. Show them what gets to grow instead. The thinking. The judgement. The conversations with clients that nobody had time for. When people can see themselves in that new picture, the workarounds gradually stop.
People is one of the five dimensions the diagnostic covers, alongside Mandate, Process, Infrastructure, and Governance.
The people quietly carrying on the old way aren't the problem. They're a signal that the rollout hasn't yet told them what they're being rolled out towards. That's a leadership job, not a training job.