There’s no shortage of conversation about AI in the workplace. Most people aren’t resisting it. In fact, many are open to it and curious about how it might help them.

But beneath that positivity, there’s a gap that’s becoming harder to ignore. New research from Strive Training shows a clear disconnect between how people feel about AI and how confident they are actually using it in their day-to-day roles.

The Confidence Gap Is Holding People Back

Across a survey of 2,000 UK employees, 64% say they feel positive about AI. Yet only 25% say they fully understand how to use it effectively in their role.

That gap matters more than it might seem. Willingness on its own doesn’t lead to action. If someone isn’t confident, they hesitate. They use AI in limited ways, second-guess their outputs, or avoid it altogether.

This is where progress slows.

People Are Trying — But They Haven’t Been Shown How

Among adult learners, nearly two-thirds say they already use AI at least occasionally. At the same time, almost a third report low confidence, despite over 75% saying they want to learn more.

This isn’t a lack of motivation. It’s a lack of clarity.

People are experimenting, but without a clear sense of what “good” looks like. They haven’t been shown how to apply AI properly in real situations, so they remain unsure whether they’re using it correctly or effectively.

Much of the Training Isn’t Working

A large part of the issue comes down to how AI is being taught. 40% of employees report receiving no AI training at all. Of those who have, only 17% describe it as highly engaging and practical.

That points to a deeper problem. Too much training is still theoretical. It explains concepts but doesn’t build capability. It introduces tools without showing how they fit into real work.

As a result, people leave knowing more about AI, but not feeling able to use it. That’s the gap.

The Divide Is Starting to Widen

The impact isn’t evenly spread. Older workers are significantly more likely to have received no training, and employees outside London are less likely to have access to useful, practical support.

Left unchecked, this creates a divide within the workforce. Some people build confidence quickly and move forward. Others fall behind, not because they can’t learn, but because they haven’t been given the right opportunity.

What Actually Builds Confidence

The research points to a clear answer. Confidence doesn’t come from more information. It comes from practice.

Learners consistently prefer hands-on, guided training built around real tasks. When they can apply AI to something familiar, such as improving a CV or organising work, the value becomes clear and confidence increases quickly.

In Strive Training’s pilot AI Skills programme, 91% of participants rated the course as Excellent or Very Good. Confidence scores increased from 5.2 out of 10 before the course to 8.8 afterwards. That shift comes from doing, not just learning.

This Is the Risk Employers Are Missing

Many organisations are already encouraging employees to use AI, and in some cases expecting it. But expectation without support creates uncertainty.

Employees are left to figure things out on their own, which leads to inconsistent use, concerns about accuracy, and hesitation around what is appropriate. Over time, this creates a split between those who feel confident using AI and those who don’t.

That gap doesn’t stay small for long.

The Shift That Needs to Happen

Closing this gap doesn’t require more content or more tools. It requires a change in approach.

Training needs to be practical, applied, and closely linked to real work. It needs to build confidence early, with clear examples of how AI can be used effectively in everyday tasks.

When people see that, they engage. When they don’t, they disengage.

The Bottom Line

AI adoption is moving quickly, but confidence is not keeping pace. That creates a risk for both individuals and organisations. If this gap isn’t addressed, parts of the workforce will be left behind, not through lack of willingness, but through lack of support.

The difference is often simple. People don’t need to be convinced that AI matters. They need to be shown, clearly and practically, how to use it.