Most CM leaders have a project on their desk they didn't ask for. The organization's AI rollout.
In a consulting firm, it's the firm's own adoption plus every client engagement that now has an AI workstream. In-house, it's the Copilot rollout, the internal AI platform, or the pilot spree the CIO is trying to coordinate.
Most are running it as a technology deployment with training attached. That's the mistake.
The mistake in one sentence.
AI adoption is a technology deployment. That's the ground truth. What makes it different is that the technology reshapes what competence looks like, and no previous deployment has done that at this scale.
Every knowledge worker built their professional identity around a set of cognitive skills. Analysis. Writing. Synthesis. Judgment. Speed. The tools they used were the visible evidence of invisible competence. That's how their manager evaluated them, and how they evaluated themselves.
The analyst proud of building a financial model in eight hours now watches a colleague produce a first draft in twenty minutes. The consultant known for the depth of his stakeholder assessments sees a junior produce something structurally similar in an afternoon. What justifies your seniority, your rate, your role has changed.
That's an identity change. Identity changes are not addressed by rolling out a system and running training.
The correct diagnosis.
Every CM methodology already has language for identity-level change. Prosci® ADKAR® at the deep end of Desire and Knowledge. Kotter's guiding coalition work. ACMP Standards' engagement work at the individual and group level. The leader running AI adoption as a system deployment isn't using them.
The correct diagnosis is a workforce-scale identity change, with every stakeholder segment asking a different version of the same three questions.
What am I losing?
What am I being asked to learn?
What can I still count on?
The analyst asks those differently from the consultant, who asks differently from the director, who asks differently from the sponsor who commissioned the rollout. Each segment needs a Three-Box view built for their reality.
That's not a training curriculum. It's a stakeholder engagement program run at scale, with middle managers doing the forgotten layer of leadership work.
So what changes?
Training becomes one workstream, not the vehicle. Impact translation gets built for every knowledge worker segment. Middle managers get engaged early and given language for the conversations their teams need. And specialist AI encoded with CM judgment does the production work, so the hours go to people, not templates.
Run it as a system deployment, get average results, move to the next project.
Run it as an identity change, become the person a leader calls before the next transformation.
Something is changing.