How Do I Get My Employees to Use AI Tools?

What stops employees from using AI at work?

Five reasons are worth checking first, and each one points at something a leader can change.

Nobody has said what it is for

As of May 2026, Gallup found 25% of US employees say their organization has communicated a clear plan for integrating AI into current practices. Without a stated purpose, people cannot tell a good use from a waste of time, so many choose not to risk it.

There is no time or skill to use it well

In BCG's 2026 survey, 72% of workers said the skills expected of them have shifted, while only 36% felt they had received adequate upskilling (BCG, June 2026). Someone who tries a tool once, gets a mediocre answer, and has no time to try again will reasonably stop.

Their manager does not use it

Managers set the norm for a team. Gallup found that, as of May 2026, 36% of employees in organizations integrating AI strongly agree their manager supports their team's use of AI (Gallup). A manager who never mentions the tool, or who still asks for the report built the old way, tells the team the change is optional.

They are worried about what it means for their job

About half of US workers (52%) told Pew they are worried about the future impact of AI use in the workplace, in a survey fielded in October 2024 (Pew Research Center, February 2025). People fill a silence with the worst explanation available, and someone who fears for their role does the minimum that is measured.

The workflow was never changed

A tool bolted onto an old process becomes an extra step. If the approval chain, the handoffs and the report format all stay the same, the AI version is more work, and people route around it to hit the deadline.

For a closer look at each reason, see why employees resist AI tools, and what leaders can do about it.

Why doesn't AI training change how people work?

Training teaches people how to use the tool. It does not change how they are measured, how their manager works, or what they think happens if the AI gets something wrong. If those stay the same, people learn the tool and then go back to the old way, because the old way is still what the system rewards.

Picture an automation platform launched across a professional services firm. Training completion is high and active use stays low.

The tool routes approvals by job title, but in that firm an approval carries weight because the right senior partner saw it. People route work around the tool to protect the informal process that works. The fix is in the approval logic, and more training would not have found it.

What gets employees to use AI tools?

Work through these in order. Each one removes a reason people have for staying with the old way.

1. Start with ground truth. Before pushing harder, find out what is happening. Where is the tool used, where is it ignored, and what are people doing instead? Ask in conversation, not only by survey. The answers become your plan.
2. Name one outcome per team. Pick the measure the tool should move this quarter, such as cycle time on contract review or first-reply time in support, and write it where the team will see it.
3. Protect time for practice on real work. Set aside a small block each week and spend it on a live task, with the colleague who already does it well showing the others. BCG's 2025 survey found employees who received five or more hours of training, especially in person and with coaching, were more likely to become regular users (BCG, June 2025). That is an association; it does not prove training causes use.
4. Equip managers first. Give managers the purpose, the outcome and honest answers about roles before their teams hear anything. Ask them to use the tool visibly in team meetings. This is the missing rung in most launches.
5. Redesign one workflow end to end. Remove the steps the tool makes redundant, retire the old report, and connect the systems people need. Adoption follows when the new way is easier than the old one.
6. Say what happens to roles and to saved time. Tell people what AI is meant to improve and where the hours it frees will go. Raise the human rather than remove the human, and say so plainly.
7. Treat resistance as information. When someone pushes back, ask what would have to change for the tool to help. Specific objections are design requirements.

Do AI mandates and usage targets work?

They raise logins, and logins are easy to mistake for progress. A target measures whether someone opened the tool, so people open it. The work stays the same, and pushback that gets filed as "change resistance" takes its information with it. Use usage data to find where to look, and judge progress by the outcome each team named.

What should happen to the time AI saves?

Decide before the launch, or the time disappears into the existing workload. In BCG's 2026 survey, 66% of frontline employees who use AI regularly said they get limited or no guidance on what to do with the time it frees (BCG, June 2026). Name where the hours go: more customer conversations, faster turnaround, work that was always deferred. That is how use turns into return.

What if employees are using other AI tools instead?

Then they have already accepted AI, and the official route is losing to the one they found. In BCG's 2025 survey, 54% of respondents said they would use AI tools even if not authorized (BCG, June 2025). Ask what problem the unapproved tool solves and why the approved one does not. That answer usually tells you what to fix in the official setup, whether that is access, speed or the systems it connects to.

How do I know if adoption is working?

Look for changed work. Ask people what they stopped doing and what they started doing, and track the outcome each team named. Logins tell you the tool was opened. For a full method, see how to measure AI adoption and how to measure the ROI of your AI spend.

Where this sits in the AI Profitability Gap

The AI Profitability Gap™ is the distance between what a company expected AI to deliver and what it is delivering. Getting employees to use AI tools covers the Usage link and two of the three conditions that sit beyond it: Direction (has anyone said what the work is for) and Skill (can people get more from the tools than the obvious). The third, Reinvestment, decides whether the time people save reaches the business.

To see which of the three is holding your teams back, take the AI Profit Readiness Assessment: eight questions, about two minutes. If you want help redesigning the launch, the AI Profit Sprint is built for it, and you can book a discovery call.

Related reading: Why do AI projects and pilots fail? · We bought Copilot and nobody uses it. What should we do?