AI Is Spreading at Work – But Plenty of Employees Are Still Sitting It Out

AP, Axios, and Gallup contributed to this report.

Artificial intelligence is creeping into more workplaces across the US But a lot of employees are still keeping their distance.

A new poll from Gallup shows usage is climbing – slowly. Around 3 in 10 workers now use AI regularly, whether daily or a few times a week. Another chunk dips in occasionally. And then there’s a large group – roughly half of the workforce – that barely touches it, or avoids it altogether.

That split says a lot about where things stand.

For some, AI is already part of the daily routine. It speeds things up, trims down repetitive tasks, and helps with everything from writing emails to sorting through data. About two-thirds of workers whose companies have rolled out AI tools say it’s made them more productive.

Managers, in particular, seem sold. Their work – planning, writing, analyzing – fits neatly with what AI tools can do. Nearly 7 in 10 leaders who use AI say it’s boosted their efficiency. Among rank-and-file employees, that number drops.

Still, even among those using it, there’s unease.

Take Scott Segal, a social worker in Virginia. He uses AI to track down healthcare resources for elderly patients – fast, practical help that saves time. But he’s also thinking ahead. If machines can handle more of that work, where does that leave him?

“I’m planning ahead,” he said.

Backup plans are already in motion.

Others are more skeptical from the start.

Even when companies make AI available, many employees just… don’t bother. Nearly half of non-users say they prefer doing things their own way. No algorithms, no prompts. Just a habit and maybe a bit of pride in skills they’ve built over years.

There are deeper concerns, too. Data privacy comes up a lot. So do ethical questions. Some workers simply don’t trust the technology, or don’t see how it fits into their job.

And then there’s the reliability issue.

Elizabeth Bloch, a labor attorney in Louisiana, uses ChatGPT to draft emails, especially in tense situations. It helps keep things measured. But when it comes to legal research, she’s cautious. AI tools can “hallucinate” – confidently producing information that isn’t real. In her field, that’s not a small mistake; it can land lawyers in serious trouble.

That hesitation shows up in the data. About a quarter of non-users say they’ve tried AI and didn’t find it useful. Others say they don’t feel equipped to use it properly.

Even among casual users, there’s a sense of picking and choosing. Thuy Pisone, a contract administrator, uses AI for routine tasks but skips it for work she already knows how to do well – like building presentations. That’s her territory.

“I don’t need help,” she said.

It took time to get good at it.

Running parallel to all this is a quieter but growing worry: job security.

The poll found that 18% of US workers think their job could disappear within five years because of AI or automation. That’s up from the year before. At companies already using AI, the anxiety is stronger.

Outside polling echoes the same mood. Many Americans expect AI to eliminate more jobs than it creates, at least in the near term.

Back in Virginia, Segal is already thinking about what comes next if that happens. His idea: a hands-on healthcare service, physically guiding patients between appointments. Something human. Something harder to automate – at least for now.

Because that’s the tension running through workplaces at the moment.

AI is clearly gaining ground. Companies are investing, managers are pushing, and productivity gains are real. But adoption isn’t automatic. It depends on trust, usefulness, and whether people feel the technology actually belongs in their day-to-day work.

Right now, a lot of them aren’t convinced.