The rollout happened. The tool is live. The vendor signed off. The IT ticket is closed.
And 3 weeks later, your team is working around it.
Not because the tool is bad. Because the tool arrived without them.
Why do employees resist AI tools after a rollout?
Employees resist AI tools not because of the technology, but because of how the change was managed. When people are excluded from decisions that affect their work, they find ways to work around the outcome.
According to BCG, only 51% of frontline workers use AI, compared to 75% of those in leadership positions. Is this a technology gap?
No, it’s a change management gap. Research reported in CNBC states that just 38% of organizations provide AI training. And only 33% of workers have recently been given any kind of AI training by their organization.
When people feel excluded from decisions that affect their work, they find ways to work around the outcome. The tool is live. The team isn’t using it. The organization has all the cost of adoption and none of the benefit.
What is the business cost of low AI adoption?
Low adoption creates wasted investment, productivity loss from managing 2 systems simultaneously, and inconsistent output — because some employees use the tool and some don’t, producing variable results for customers.
The business cost is significant:
- Wasted investment. Enterprise AI tools are not cheap. A tool sitting unused or underused represents direct cost with no return.
- Productivity loss. Employees managing 2 systems — the new AI tool and their established workarounds — are slower than employees using either one alone.
- Inconsistent output. When some employees use the tool and some don’t, the organization produces variable results. Some customers get the AI-assisted experience. Some don’t. The variance itself becomes a quality problem.
| Adoption approach | Outcome |
|---|---|
| Tech-focused rollout (tool first, people second) | 1.6x more likely to miss investment-return expectations (Deloitte) |
| Human-centric rollout (people + training + leadership support) | Positivity about GenAI increases from 15% to 55% (BCG) |
| No formal AI training provided | Only 24% of employees feel equipped to use AI (Skillsoft) |
How does low employee AI adoption affect customers?
Employees who don’t trust the AI tool are slower and less consistent. Customers experience this directly — longer wait times, variable answers, and a service that feels less reliable than before the ‘upgrade.’
There’s a direct line between how employees experience an AI rollout and how customers experience the company.
An employee who doesn’t trust the tool is slower. Slower response times mean longer customer wait times. An employee who doesn’t know when to trust the AI output gives inconsistent answers. Inconsistent answers erode customer trust.
According to Deloitte, organizations taking a tech-focused approach to AI implementation are 1.6 times more likely to fail to exceed investment-return expectations than peers taking a human-centric approach — confirming that the people side of implementation is not secondary to the technical side.
What does human-centered AI implementation look like in practice?
Human-centered implementation involves employees in tool selection before the choice is made, builds training into the implementation timeline as a requirement, defines adoption metrics before launch, and creates a feedback loop for ongoing improvement.
In practice, human-centered AI implementation means:
- Involve employees in tool selection. Front-line employees understand workflow constraints that leadership often doesn’t. Including them in the evaluation process surfaces practical objections early, when they’re cheap to address.
- Build training into the implementation timeline as a requirement. Not a launch-day webinar — a structured program that runs before the tool is live in production.
- Define adoption metrics before launch. What does successful adoption look like at 30 days? 90 days? Set the target before launch.
- Create a feedback loop. Employees who use the tool daily will identify failure modes that testing and QA miss. A lightweight ongoing feedback mechanism captures this intelligence.
- Communicate the why, not just the what. Employees who understand why the organization is adopting AI — and specifically how it’s intended to make their work better — are significantly more likely to engage with the change.
Frequently asked questions
Quick answers to the questions leaders most commonly ask about this topic.
Why is employee AI adoption so low at many companies?
Research consistently points to change management failures, not technology failures. Employees aren’t prepared before tools go live, aren’t involved in tool selection, and don’t receive adequate training.
How does employee AI adoption affect the customer experience?
Directly. Employees who don’t trust or use the AI tool produce inconsistent, slower outputs. Customers experience this as longer wait times, variable answers, and degraded service quality.
What is human-centered AI implementation?
A rollout approach that involves employees in tool selection before the decision is made, provides structured training before go-live, defines adoption metrics in advance, and creates a feedback mechanism for ongoing improvement.
How much AI training do employees typically receive?
Very little. Only 33% of workers have received AI training from their companies, and only 38% of companies offer AI training at all.
This is the third article in the 4-part ‘Is your AI strategy actually a strategy?’ series. Read the rest:
Article 1: Your team is already using AI. That’s not the problem.
Article 2: The strategy was fine. The handoff killed it.
Article 4: What a real AI strategy looks like for a company your size.
