More than one in four U.S. workers now fear technology will make their job obsolete

The share of American workers who worry that technology could soon make their jobs obsolete has reached a new high. In polling released in mid-September, Gallup found that 27% of U.S. workers hold that concern, seven percentage points more than a year earlier and roughly double the 13% the firm recorded when it began asking in 2017.

The more telling detail is who is driving the increase. In 2017 and 2018, workers in the first and second halves of their careers expressed similar levels of worry. Now the gap is 15 points: 34% of workers aged 18 to 44 are concerned, against 19% of those 45 and older. For the younger group, technology displacement has become one of the most common job worries Gallup measures.

College graduates, long assumed to be the group most insulated from automation, have also moved: 25% now say they are worried, up six points from last year. Meanwhile, concern about the traditional threats Gallup tracks alongside this one, such as layoffs, reduced hours, lower wages and benefit cuts, has stayed broadly flat. That matters, because it suggests the rise is not just general economic nerves wearing a new label. It is specific to technology.

The fear is running ahead of the data, for now

Gallup itself is careful on this point. Its analysis describes AI’s overall effect on U.S. employment so far as modest, while noting signs of weaker hiring for entry-level and junior roles in occupations more exposed to AI. That is consistent with the pattern a lot of employers describe privately: fewer new junior hires in some functions rather than mass layoffs of existing staff.

Workers do not experience that nuance. They experience the job posting that did not open, the internship program that shrank, and a steady drumbeat of headlines. For younger workers early in their careers, a slowdown in entry-level hiring is not an abstraction. It is the ladder they were planning to climb.

Why this is an employer problem, not just a news story

Anxiety has costs that show up long before any automation does. Workers who believe their role is disappearing are more likely to start looking elsewhere, less likely to volunteer for new tools, and quicker to read every process change as a threat. The irony is that the employees most worried about being replaced are often the ones a company most needs to adopt new tools well.

Small and mid-sized businesses are in a particular position here. They rarely have formal change-management programs, and an owner’s offhand comment about “letting software handle that” can travel through a 20-person company in an afternoon.

What employers can do this quarter

1. Say what you are actually using AI for. Name the tasks, not the ambition. “We use it to draft first replies to routine customer emails, which a person reviews” is reassuring in a way that “we’re going all in on AI” is not.
2. Put the people doing the work in charge of the pilot. The fastest way to turn a threatened employee into an advocate is to make them the one who decides where the tool helps and where it gets in the way.
3. Protect the entry-level pipeline on purpose. If AI absorbs some junior tasks, decide deliberately how new people will still learn the trade. Businesses that stop training juniors today will be short of seniors in five years.
4. Measure time saved and say where it went. If a tool frees ten hours a week, tell the team whether those hours went to better service, new work or shorter weeks. Silence gets filled with the worst interpretation.
5. Write a one-page AI-use policy. What tools are allowed, what data never goes into them, who reviews output, and how customers are told. Clarity lowers the temperature.

For workers reading this

The same data points to a practical response: familiarity tends to reduce both risk and fear. Workers who use the tools in their own role are better placed to show where human judgment still matters, and to move into the work that grows around the tools rather than the work being absorbed by them.