Look at your AI rollout as a change management project
At LinkedIn’s Talent Connect conference last month, CEO Dan Shapero said every job is going to be a new job. If that is true, most organizations are preparing for it with a training module and
a completion deadline.
But training is rarely where adoption stalls. Andus Labs, an AI readiness consultancy, tracks what stalls enterprise AI. In its latest Ground Truth Index, 12 of the top 25 barriers trace back to leadership. Most are problems HR already knows how to work on: workload, incentives, job design, and the order in which change happens.
Three patterns stand out
Learning while drowning. Companies add mandatory AI training on top of full workloads, then read low completion as resistance. Andus Labs CEO Chris Perry puts it plainly: “A training mandate is a workload decision.” If nothing comes off the plate, employees choose what to neglect. A finished module also tells you very little about whether someone can use the tool on their real work.
The one-sided deal. Many rollouts explain what the company expects from AI and leave out what employees get in return. If someone gets faster, what improves for them? Better work, a path to advancement, time back, or simply more work? Without an answer, people are being asked to commit to a change with no defined benefit to them.
Assumed obsolescence. When employees connect AI efficiency with job cuts, they complete the required training and keep their best ideas to themselves. Perry calls it “visible compliance, private withdrawal.”
I have lived a version of this outside of AI. On a technology rollout I manage in healthcare education, the platform arrived at the same time hospitals began enforcing requirements that had been loosely followed before. People experienced the platform as the burden. The requirements were the change, and the tool took the blame.
AI rollouts carry the same risk. If the real change is a new expectation for how work gets done, name that change on its own. Otherwise, the tool absorbs every frustration.
Where HR can help:
Clear capacity before setting a deadline. Decide what people will stop doing while they learn.
Teach at the point of need. Training sticks when people can see where it applies to their own work. A front-loaded session fades by week three.
Write the employee side of the deal into the plan. Say how the organization will use the time AI creates.
Measure use. Completion rates show attendance. Watch whether people are using the tools on real work.
Every job may become a new job. People will learn the new one faster when someone has made room for them to learn it and told them what is in it for them.