For the first time in Gallup’s measurement, half of employed American adults now use AI in their jobs at least a few times a year. That is the headline from a new Gallup survey of 23,717 U.S. employees conducted in February 2026. Usage is up from 46% just one quarter ago.

It sounds like a win. And in some ways it is. Crossing 50% in a technology that barely registered as a workplace tool two years ago is a real shift.

But the other number in the report is the one business leaders should be thinking about.

Forty-nine percent of workers say they never use AI in their role. Not infrequently. Never.

The Real Story Is the Gap

The split is not random. Gallup’s data shows that AI adoption in the workplace tracks closely with three factors: whether the organization has rolled out AI tools, whether managers actively encourage use, and whether employees feel they have the skills to use AI effectively.

On the organizational side, 41% of employees say their company has integrated AI technology or tools to improve organisational practices, up three points from the previous quarter. That means 59% of workers are in organisations where no formal AI integration has happened yet.

This is the adoption gap. It is not primarily a technology problem. It is a training and culture problem.

The workers who are using AI regularly (13% daily, 28% several times a week) are clustered in organisations where the technology has been deployed with intent, where training was provided, and where using AI is normalised. Everyone else is either waiting to be shown how or actively choosing not to engage.

Why Some Workers Are Opting Out

Gallup’s research also captures why the 49% non-users are staying away. The main reasons are not surprising:

• Preference to work without AI
• Ethical concerns about the technology
• Data privacy worries
• Uncertainty about whether it is actually useful for their specific work

These are not irrational objections. They are the natural response to being given access to a tool without adequate context, support, or training. Workers who have not seen AI produce real results for their specific tasks are going to remain sceptical.

The 18% of workers who say it is likely their job will be eliminated by AI in the next five years, up from 15% in 2025, are navigating a genuine fear. But anxiety about job displacement does not automatically translate into AI adoption. For many workers, it produces the opposite: disengagement from the technology.

What This Means for Organisations Right Now

The adoption gap has a direct business cost. Organisations where AI use is clustered among a minority of tech-forward employees are not getting the productivity returns they should be. Competitive advantage from AI does not come from having access to the tools. Every company has access to roughly the same tools. It comes from the breadth and depth of deployment across teams.

If your finance team is running AI-assisted analysis but your operations team is doing everything manually, you have an uneven playing field inside your own organisation. The gap compounds over time as the AI-using teams get faster, iterate on what works, and build skills, while the non-using teams fall further behind.

Gartner has projected that 40% of enterprise applications will include task-specific AI agents by the end of 2026. The organisations that will benefit most from that shift are the ones where employees are already comfortable working alongside AI tools, where the cultural groundwork has been laid.

The Gallup data suggests that groundwork is still missing in most workplaces.

What This Means for Business

The 50% milestone is a marker that AI adoption has crossed from niche to normal. But “normal” is not the same as “effective.” Half the workforce using AI occasionally is very different from half the workforce using AI skillfully and consistently.

The businesses that will pull ahead over the next 12 months are the ones that close the gap between tool access and actual proficiency. That requires intentional upskilling. Not just deploying tools and hoping people figure it out, but building structured learning pathways that meet employees where they are.

For data and AI skills specifically, that means starting with the tools people already use (spreadsheets, reporting tools, dashboards) and building toward AI fluency from a foundation that feels familiar.

The 49% who are not yet using AI are not a problem to be solved with a policy mandate. They are an opportunity waiting for the right training investment.

---

If your organisation is working through the challenge of AI adoption at scale, EDNA Learn provides structured data and AI training programs built for business teams, from foundational data literacy through to advanced AI application. The gap between AI access and AI proficiency is a solvable problem.