The AI Anxiety Economy: Why 61% of Workers Now Fear the Axe

A new national survey finds automation anxiety doubling, workers willing to trade pay for job security, and widespread distrust of corporate ‘AI layoff’ claims.

Hard Reset is going to Sundance. Join us on Monday, January 26th at the film festival for a conversation about monopoly power in the streaming era and whether Hollywood’s creative labor can survive it. Register here! [Event sponsored by Parch non alcoholic drinks]

Here’s how relaxed people were about the job market just seven years ago: At the end of 2019, 3.5 million Americans were voluntarily quitting their jobs each month. Gainfully employed Americans were walking right out the door just before the holidays, presumably looking for a little down time and confident that they could find another, better gig in the new year. That “quits” rate was 2.3 percent, the highest since the government started tracking the number in 2000.

That was then. Today, 61% of employed Americans now report “layoff anxiety” — up nearly 30% since 2019, according to a new report commissioned by a global HR firm. And in the minds of an increasing number of them, A.I. is the threat. A Harris poll for Los Angeles-based INTOO reveals that Americans at every level of experience — but especially midcareer employees — are not only worried that they’re going to get the axe, they’re actually willing to take less money just to be able to relax a little. According to the report’s authors, “more than half of employees (52%) say they would take a pay cut for a job that is ‘layoff-proof’ for two years.”

In 2019, fully 3.5 million Americans were voluntarily leaving their jobs each month. Today, more than half of them are worried about being laid off. Photo by Jakub Żerdzicki on Unsplash

The new desire for stability — even at a measurable personal price — is a major generational shift, as careers that in the past defined stability are suddenly targeted for big layoffs. Last month Wells Fargo cut 5,600 people, with its CEO specifically citing AI as making the bank able to get by with fewer workers from now on.

Wells Fargo is the first major American bank to describe A.I. layoffs so frankly, but they’re following the world’s lead. DBS, a large Asian bank, announced in February of 2025 that it would eventually cull 4,000 workers. And as early as October of 2024, Italian bank Intesa Sanpaolo announced the end of 9,000 jobs. Both banks, like Wells Fargo, cited A.I. in their announcements.

But the Harris poll suggests that while increasing numbers of Americans worry about the threat of automation, they’re not exactly convinced that what companies are telling them is true. They’re skeptical that layoffs are necessary, and they think A.I. is a smokescreen. 76% of employees told Harris that most layoffs in recent years are a result of bad business management. And “there is also deep skepticism about whether the current and future role of AI at work is driving workforce-reduction decisions,” the authors write. “69% of employees agree that ‘AI layoffs’ are just an excuse for cost-cutting.”

So where should Americans look for the security they crave? Another recent study asked Americans which companies inspired their trust in the face of A.I. layoffs. First on the list was Vermont’s Ben & Jerry’s, followed by Georgia’s Coca-Cola and Hawaii’s Four Seasons Hotels and Resorts.

The overall sense from respondents was that healthcare is the safest bet, with Illinois’s Blue Cross Blue Shield, Maryland’s Johns Hopkins, and Florida’s Advent Health just behind the top three. But that industry is also dropping knowledge workers, with 51 hospitals and healthcare systems announcing layoffs that will affect tens of thousands of Americans. And while those cuts haven’t been directly attributed to A.I., an unusual number of the announced layoffs affect administrative staff, suggesting a streamlining of office functions that goes hand-in-hand with the efficiency promises of the technology.

It’s not clear when workers can relax again. For one thing, no corner of the United States seems safe. Virginia’s chamber of commerce predicts that 500,000 jobs are under threat in that state alone, and a November 2025 study led by MIT says that workers’ exposure to AI-related job loss is so clearly and evenly distributed nationally that we’ll need an entirely new way of measuring who is at risk.

Downsizing may be a fundamental downside of the technology that the job market can’t solve. The A.I. industry has said that its technology can usher in breathtaking advancements, but that they may come with breathtaking unemployment. Dario Amodei, cofounder of Anthropic, warned in a 2025 Axios interview that A.I. may pull us into a world where “cancer is cured, the economy grows at 10% a year, the budget is balanced — and 20% of people don't have jobs.”

At the moment, it’s a rare CEO that will openly admit that A.I. is the company’s rationale for letting its people go. But that could change. The AI-Related Job Impacts Clarity Act, introduced in November by Senator Josh Hawley (R-MO), would require that large employers and federal agencies tell the Bureau of Labor Statistics each quarter how many people they’re laying off largely because A.I. can do the job instead, how many jobs are going unfilled because A.I. replaced the humans involved, and the number of workers retrained because A.I. made it necessary. Not a fix for the problem, but at least a way of measuring it.

In the meantime, the Harris report’s authors write, this doesn’t appear to be a temporary fluctuation in worker confidence. As the economy slows, and A.I. accelerates, “rising concerns tied to the broader economy, automation, and recession expectations signal that employees anticipate instability to persist.” The breezy job quitters of the late 2010s may be the last Americans in our lifetime to find a better employer so easily, unless a fundamental shift — in the way labor is organized or regulations are written — pushes us back that way.

What we’re reading

We linked to this sign on letter last week, this week Wired reported about it: Tech Workers are Condemning ICE, even as their CEOs Stay Quiet

404 Media provided a very visual description of how ICE agents use Palantir’s software to find people.

The WSJ reported on ICE officers apprehending construction workers building a Meta data center in rural Louisiana.

Popular Information reported on the companies still advertising on X this week as Grok produced sexualized images of kids

New research found datacenters were one of the major drivers of the first rise in total economy-wide emissions in the US in years, via Ketan Joshi on Bluesky