Some of the biggest names in AI warned that unemployment could spike as high as 20%...
In 2025, Anthropic CEO Dario Amodei said the technology could wipe out half of all entry-level white-collar jobs within five years.
And the news spread fast. A record 27% of U.S. workers fear that technology could soon eliminate their jobs, according to analytics firm Gallup.
Then the official data came out...
The U.S. unemployment rate currently sits at 4.1%. That's about one-fifth of the worst-case scenario (20%). In August, employers added 162,000 jobs, roughly triple Wall Street's forecast of 53,000.
Across the Atlantic, a new survey suggests that AI is starting to add jobs.
U.K. banking giant Lloyds Banking (LYG) polls about 1,200 British businesses every month. In its latest survey, 54% of them said AI had created new jobs.
The fear makes sense, though. Folks have watched AI write code, answer customer calls, and draft legal briefs faster than its human counterparts. It's easy to imagine a wave of layoffs tipping the economy into recession.
But as we'll explain, the evidence shows that AI is reshaping work far more than it's erasing it.
Those most exposed to AI tend to be college-educated office workers...
If AI were gutting white-collar work, they'd feel it first. But they aren't.
The unemployment rate for workers with a college degree is 2.7%. It has held at that level since May.
U.S. employers blamed about 55,000 layoffs on AI in all of 2025. That's less than 5% of the year's job cuts.
And nearly four years into the AI boom, the unemployment rate hasn't budged very much.
Take a look...
Since ChatGPT kicked off this tech surge in late 2022, the unemployment rate has stayed between 3.4% and 4.5%.
But AI isn't harmless.
Financial powerhouse Goldman Sachs (GS) estimates that AI has trimmed U.S. payroll growth by about 16,000 jobs per month. That has added roughly one-tenth of a percentage point to the unemployment rate.
It's a real loss for the people affected. But in an economy with nearly 160 million jobs, it's a drop in the bucket. And it's far from a recession trigger. (It's important to note that the next jobs report comes out tomorrow, October 2.)
Businesses are building hiring practices around AI technology...
And the recent Lloyds survey shows us how.
One-quarter of U.K. firms are hiring more candidates with AI skills. One-fifth of them are creating brand-new AI roles. And more than 70% of large businesses plan to invest more money into the technology.
Amanda Murphy, CEO of Lloyds' business and commercial banking division, told Bloomberg that firms need to focus on "building the skills, culture and confidence to use [AI] effectively."
Some companies have learned that the hard way.
"Buy now, pay later" company Klarna (KLAR) froze hiring because its AI assistant could do the work of 700 customer service agents. More than a year later, CEO Sebastian Siemiatkowski started hiring people again. Chasing cost savings had diminished the quality of Klarna's support services.
This business pattern has played out before...
When U.S. banks rolled out loads of ATMs in the 1990s, tellers looked doomed. ATMs handed out cash and took deposits, the core of a teller's job.
But the tellers stuck around.
Between 1988 and 2004, ATMs cut the number of tellers in the average U.S. city branch from 20 to 13. That made each branch cheaper to run... so banks opened more.
Total teller jobs held steady and even rose. The work shifted toward selling banking services and solving tough customer problems.
Goldman sees the same split happening with AI.
In cases where AI can fully replace a worker, jobs are falling. But where AI helps workers do more, employment is rising. The technology is adding about 9,000 jobs per month.
It's another version of the teller story.
Don't bet your portfolio on an AI recession...
AI will cost some people their jobs. It already has.
But the data shows no sign of a mass-unemployment event that would drag the U.S. into a recession. Businesses are hiring folks with in-demand AI skills. And they're creating new roles, just as banks did when ATMs arrived.
Investors shouldn't sell stocks just because a tech CEO predicts a jobs apocalypse.
Track the monthly jobs numbers instead, starting with tomorrow's report. As long as unemployment holds near 4.1%, the AI-recession story is just that... a story.
And when you size up a company's AI plans, look at how it actually uses the technology. Staff cuts give profits a one-time bump. Using AI to serve more customers and open the next company "branch" can fuel growth for years.
Those are the businesses worth owning.
Regards,
Rob Spivey
October 1, 2026
