America's CAR-MART

Most organizations are moving quickly on AI.

Leaders announce the strategy. New tools are introduced. Licenses are distributed. HR sets guardrails. Training is launched. Employees are encouraged to experiment.

Then several weeks pass, and much of the daily work looks the same. I’ve seen how quickly that gap between expectation and reality can frustrate leaders and employees alike.

People may use AI to summarize a document, draft an email, or organize information. Yet the way the team works, solves problems, makes decisions, or serves customers has changed very little.

That is where many AI efforts begin to stall.

Where Strategy Meets the Work

A company can provide the tools, set expectations, and train employees. Someone still has to answer the questions that follow: Where does this help our team? When should we use it? What should still depend on our own judgment? What happens when AI gets it wrong?

For most employees, those answers come from their manager. That is the missing middle of AI transformation.

Implementation can be centralized. Adoption is local.

Senior leaders can explain why AI matters. Managers help employees understand what it means for their work.

Telling a team to “use AI” may create activity. Helping people identify where it can improve the work creates value.

Managers are close enough to see where time is lost, work is repeated, or better preparation could improve a decision or customer interaction. They can look at a real process and ask: Is there a better way to do this?

That is how AI starts becoming part of the work itself.

Trust Shapes Adoption

Employees also need to feel comfortable using it openly.

AI raises questions people may never say out loud. Will using it make me look less capable? Is it considered taking a shortcut? What happens if I trust an answer that turns out to be wrong?

A policy can establish the boundaries. Managers establish what using AI actually feels like on the team.

Employees pay attention to what happens when someone experiments. They notice how a manager responds when an AI-assisted result misses the mark and whether concerns about a questionable answer are encouraged or dismissed. Those moments reveal whether experimentation is truly supported.

They also build capability.

“Telling a team to “use AI” may create activity. Helping people identify where it can improve the work creates value.”

Employees get better with AI by applying it to real work, seeing where it helps, recognizing where it falls short, and adjusting their approach. Training can provide the foundation, but capability develops through repeated use.

Managers strengthen that learning by asking better questions, challenging weak conclusions, encouraging employees to verify what AI produces, and reinforcing where human judgment matters most.

The Manager Cannot Be an Afterthought

There is a problem with placing so much responsibility on managers: they are already carrying a significant load.

They are responsible for results, staffing, coaching, development, administration, and the steady stream of changes organizations ask teams to absorb. Now they are also being asked to help employees figure out how AI fits into the work.

Adding “drive AI adoption” to an already crowded list of expectations will not move an organization very far.

Managers need to be prepared before the rollout reaches their teams. They need firsthand experience with the tools, clear guidance on where experimentation is encouraged and where human judgment must remain central, and time to learn before they are expected to coach others.

Build the Rollout around the People Who Make It Real

Too often, managers receive the same announcement, training, and access as everyone else, then are expected to answer the harder questions once the rollout begins.

That sequence needs to change.

Managers should be among the first people prepared. Before employees are asked to change how they work, their managers should understand where AI can help, what responsible use looks like, what the organization expects, and how to coach through mistakes and uncertainty.

I’ve seen firsthand that employees rarely experience transformation the way a strategy deck describes it. They experience it through conversations, priorities, feedback, and everyday decisions made by the person leading their team.

When those daily moments reinforce the strategy, AI begins to change the work. When they do not, the rollout becomes another initiative employees heard about, tried briefly, and eventually worked around.

The organizations that get this right will have managers ready to help people use AI well, question it when necessary, and improve the work around it.

That is the point where AI stops being something the organization has and starts becoming something the organization knows how to use.

AI strategy may be set at the top.

AI adoption is built in the middle.